Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Jun 1.
Published in final edited form as: Am J Prev Med. 2021 Mar 27;60(6):866–872. doi: 10.1016/j.amepre.2021.01.012

Validity of the Exercise Vital Sign Tool to Assess Physical Activity

Jennifer L Kuntz 1, Deborah R Young 2, Brian E Saelens 3, Lawrence D Frank 4,5, Richard T Meenan 1, John F Dickerson 1, Erin Keast 1, Stephen P Fortmann 1
PMCID: PMC8154650  NIHMSID: NIHMS1674227  PMID: 33781618

Abstract

Introduction:

Assessment and counseling by healthcare providers can successfully increase physical activity (PA); however, a valid instrument to effectively measure PA is needed. This study examines the validity of the Exercise Vital Sign (EVS) tool by comparing EVS data collected at Kaiser Permanente Northwest to accelerometry data.

Methods:

Participants (n=521) completed accelerometer monitoring and had ≥1 EVS measurement in their electronic medical record. Using accelerometry as the “gold standard,” the association between moderate-to-vigorous PA (MVPA) minutes/week estimated through EVS and accelerometry was examined using the Spearman correlation coefficient. Comparability of MVPA categories (inactive, low active, moderately active, sufficiently active) was assessed using simple and weighted κ statistics. Sensitivity, specificity, and positive and negative predictive value were calculated. The study was conducted in 2012–2015, with analysis in 2019–2020.

Results:

Average accelerometry-based MVPA was 212 minutes/week and 57% were considered sufficiently active. EVS-based MVPA averaged 170 minutes/week and 53% were active. There was a positive correlation between MVPA minutes/week reported through EVS and accelerometry (r =0.38, p<0.0001). Fair agreement was observed between EVS- and accelerometry-based MVPA categories (weighted κ=0.29), with the highest agreement occurring for those with ≥150 minutes/week. The positive correlation increased when MVPA was examined dichotomously (<150 or ≥150 minutes/week, κ=0.34). The sensitivity, specificity, positive predictive value, and negative predictive value for EVS (when compared with accelerometry) were 67%, 68%, 61%, and 73%, respectively.

Conclusions:

The EVS is a useful PA assessment tool that correctly identifies the majority of adults who do and do not meet PA guidelines.

INTRODUCTION

Physical activity (PA) is associated with health benefits, including reductions in risk for cardiometabolic diseases, some forms of cancer, and premature mortality.1 Despite this, in 2018, approximately one half of U.S. adults were either inactive or engaged in some PA but not enough to meet the target threshold of 150 minutes of moderate-to-vigorous PA (MVPA; e.g., intensity of brisk walking or greater) recommended by national guidelines.2 Collectively, this lack of PA is linked to approximately $117 billion in annual healthcare costs and roughly 10% of premature mortality.1

National recommendations acknowledge the role of healthcare providers and systems in addressing physical inactivity and endorse PA counseling among adults with obesity or cardiovascular risk factors.3-5 Although these formal clinical guidelines do not extend to the general adult population, several national initiatives encourage healthcare providers to assess activity levels and, when appropriate, provide education and counseling to promote PA.5-8 These recommendations are supported by research showing that brief assessment and counseling by healthcare providers can successfully increase PA.9-12 However, for healthcare providers to be able to assess PA and identify patients for counseling, they must have access to a valid assessment instrument that is also brief, easy to score, and easily integrated into clinical workflow.

In 2009, Kaiser Permanente Southern California led the healthcare sector in its effort to assess PA by implementing the Exercise Vital Sign (EVS) tool in its electronic medical record (EMR).13 The EVS tool assesses the average minutes/week spent exercising based on 2 self-reported questions. The goals of the EVS are to ensure assessment of patient PA habits at each visit and facilitate brief counseling that encourages patients to increase activity to the recommended levels of MVPA. The EVS was gradually introduced in all KP regions and was implemented in Kaiser Permanente Northwest (KPNW), the site of this study, in 2012.

The EVS has shown good face and discriminant validity.13-16 Using EVS measurement, inactive patients were older, more likely to be female, more likely to be Hispanic or Black, and had higher BMI levels and more chronic health conditions, when compared with patients who met MVPA guidelines.13 Additional research has shown that patients who were consistently active based on the EVS had lower blood pressure and glucose levels compared with patients reporting low PA levels.14 However, there is insufficient information on criterion validity of the EVS compared with accelerometer data. Available studies have been conducted among small populations and have not examined the validity of EVS data collected in real-world, clinical care settings. In 1 study, among 76 participants with 7-day accelerometry data, the EVS tool showed a moderate ability to correctly identify participants as not meeting or meeting PA guidelines. However, the study also found that the EVS significantly overestimated the number of minutes/week of MVPA, which may result in missing patients in clinical care who should be advised to increase their PA.15 An additional study conducted among 30 African American women showed a weak correlation between EVS- and accelerometry-based MVPA measurement.16

This study sought to further examine the validity of the EVS by comparing EVS data collected at KPNW during routine clinical care to accelerometry data collected among a large sample of adults participating in an existing research study. Analyses particularly focused on the validity of the EVS for identifying individuals who are inactive and insufficiently active and who may benefit from PA counseling.

METHODS

Study Population

This study was conducted at KPNW, an integrated healthcare system in which comprehensive medical care is delivered to about 600,000 members in northwest Oregon and southwest Washington. The study was conducted in 2012–2015; data analysis occurred in 2019–2020.

This study was conducted as part of the Rails & Health study, a longitudinal natural experiment examining the impact of a new light rail transit line opening in Portland, Oregon on health outcomes, healthcare utilization, and cost.17 The study population included adult KPNW members (aged 18–74 years) who lived at their residential address for ≥3 years before September 2015 (the opening of the light rail transit line). Participants were recruited to complete PA monitoring using an accelerometer for a 7-day period. To participate in PA monitoring, members were required to have no debilitating chronic conditions and be capable of PA. For inclusion in the analyses, participants were required to have ≥1 EVS measure available from the KPNW EMR and ≥6 days of valid accelerometer data. All study procedures and materials were approved by the KPNW IRB.

Measures

Demographic information including age, sex, and race/ethnicity were collected from the EMR. Age was calculated as of the first date of accelerometer data collection. Educational ascertainment was captured through a study survey. BMI was calculated using EMR height and weight measurements closest to the first date of accelerometry. Height and weight are captured in the EMR at each healthcare visit and have been validated as highly reliable.18,19 Chronic conditions were identified using diagnoses recorded in the 1 year prior to study recruitment and aggregated using the Charlson Comorbidity Index.20

All EVS measures recorded in the EMR between September 1, 2012 and September 14, 2015 were included. The EVS questions are administered by a medical assistant during vital sign measurement for all outpatient visits unless a prior measurement was collected within the last 90 days. Answers are entered directly into the EMR. The EVS assesses the average time spent exercising through 2 questions: On average, how many days per week do you engage in moderate to strenuous exercise (like a brisk walk)? and On average, how many minutes per day do you engage in exercise at this level? 13 The response to the first questions was entered as a categorical variable, with values 0–7. Minutes are recorded in blocks of 10 (0, 10, 20, 30, 40, 50, 60, 90, 120, and ≥150). The EMR software then multiplies these responses to display minutes per week of moderate or strenuous exercise.13 The analyses utilized minutes of MVPA per week as available from EMR data.

Accelerometer data were collected from December 2014 through September 2015 using the Actigraph GT3X+.21,22 This triaxial monitor detects acceleration in the vertical, anteroposterior, and mediolateral axes and is a reliable and valid device for measuring PA.23 Vertical axis (or axis 1) counts were used to derive PA intensity. Participants were asked to wear the accelerometers on a study-provided belt that fit around their waist and to position the accelerometer over their right hip during all waking hours for 7 consecutive days. Participants were also asked to record times when they were not wearing the accelerometer. The accelerometry data were collected and stored in 15-second intervals. Non-wear time was defined as time periods of ≥20 minutes with 0 counts/60-second interval. Accelerometer data were categorized into PA bouts. Bouts of PA (inclusive of light, moderate, and vigorous PA) were defined as time intervals having >500 accelerometer counts per 60-second epoch for at least 7 minutes, where only 2 minutes can be below the threshold.17 MVPA (moderate and vigorous PA only) was defined in a similar fashion but at a threshold of ≥1,952 counts/60-second interval.24

Statistical Analysis

Summary statistics were calculated for demographic and clinical characteristics. The median of all EVS values for participants during the study period were calculated, to account for differences in self-reported exercise across healthcare visits. For inclusion in the analyses, ≥6 days of valid accelerometer data were required. Accelerometer data for each day were considered valid if data were available in a participant’s record for ≥8 hours, with non-wear time removed. Only the first 7 days of “valid” data collection were included in cases where a participant wore the accelerometer for more than the 7-day data collection window. For analyses, time across PA bouts was summed to calculate MVPA minutes per week. The maximum allowable value within the EMR for MVPA based on EVS is 1,050 minutes/week (i.e., 150 minutes/day). To limit the occurrence of extreme values and potential outliers, values recorded through accelerometry were Winsorized at a threshold of 1,260 minutes/week. Winsorizing was used to limit the influence of extreme observations without removing them entirely from the analysis sample.25 As done here, the process sets values above a specified percentile (e.g., 99th percentile) to that threshold percentile value.25

The validity of the EVS measure was assessed both continuously and categorically, using accelerometry data as the “gold standard.” For categorical analyses, EVS and accelerometry data were divided into 4 categories based on data distribution and national guidelines. EVS-based MVPA was categorized as: inactive (0–9 minutes/week), low active (10–59 minutes/week), moderately active (30–149 minutes/week), and sufficiently active (≥150 minutes/week). Because accelerometry data capture short bouts of activity that a participant may not self-report, MVPA assessed through accelerometry was categorized as: inactive (0–29 minutes/week), low active (30–59 minutes/week), moderately active (60–149 minutes/week), and sufficiently active (≥150 minutes/week). The data were also examined dichotomously as <150 or ≥150 minutes/week, in an effort to examine the clinical utility and validity of the EVS in identifying individuals most in need of counseling about recommended PA levels.

The Spearman correlation coefficient was used to examine the association between minutes/week of MVPA per recorded through EVS and accelerometry. Simple and weighted Cohen’s κ statistics were used to examine agreement between categorizations of MVPA. The sensitivity, specificity, positive predictive value, and negative predictive value were calculated for the dichotomous categorization of MVPA through the EVS, as compared with accelerometry.

RESULTS

The study consented 666 participants, of which 578 completed PA monitoring. These analyses included 521 participants who further met inclusion criteria based on availability of EVS and valid accelerometry data. The mean age of participants was 55 years, 65% were female, and 90% were non-Hispanic White. On average, participants had a BMI of 27.4 and a Charlson Comorbidity Index score <1 (Table 1).

Table 1.

Demographic and Clinical Characteristics of Study Population by Categorization of Physical Activity Measured Through Accelerometry

Demographic and clinical characteristics PA categorization by accelerometrya
Inactive
n=54		 Low active
n=42		 Moderately active
n=127		 Active
n=298		 Overall
n=521
Age, mean (SD) 62.3 (11.2) 59.9 (11.3) 55.4 (13.0) 52.7 (12.0) 55.0 (12.5)
Female, N (%) 44 (81.5) 31 (73.8) 85 (66.9) 177 (59.4) 337 (64.7)
Non-Hispanic White, N (%) 49 (90.7) 34 (81.0) 115 (90.6) 274 (92.0) 473 (90.8)
Education, N (%) college graduates 24 (44.4) 23 (54.8) 90 (70.9) 223 (74.8) 360 (69.1)
BMI, mean (SD) 32.0 (7.9) 31.6 (7.9) 27.6 (5.2) 26.0 (4.2) 27.4 (5.7)
Charlson comorbidity score, mean (SD) 1.0 (1.1) 0.5 (0.9) 0.3 (0.6) 0.2 (0.6) 0.3 (0.7)
EVS, minutes/week, mean (SD) 80.9 (99.9) 109.5 (130.2) 137.7 (125.7) 208.7 (151.5) 170.2 (146.7)
Open in a new tab
a

Inactive defined as 0–29 minutes/week; low active defined as 30–59 minutes/week; moderately active defined as 60–149 minutes/week; active defined as ≥150 minutes/week.

EVS, exercise vital sign; PA, physical activity.

On average, participants wore the accelerometer for 7 days and had a mean wear time/day of 12 (IQR=11–14) hours. Participants engaged in an average MVPA duration of 212 (SD=169, median=183, IQR=87–289) minutes/week based on accelerometry data. Participants had an average of 7.3 (SD=6.2, median=6, IQR=3–10) EVS measurements in the EMR. The average weekly MVPA based on EVS was 170 (SD=147, median=150, IQR=65–240) minutes/week.

Based on accelerometry data, 57% of participants met the criteria for being active, while 24%, 8%, and 11% of participants were moderately active, low active, and inactive, respectively (Table 1). Based on EVS data, 53% of participants reported ≥150 minutes/week and met the criterion for being sufficiently active, while 26%, 5%, and 16% of participants were categorized as being moderately active, low active, and inactive. Participants defined as active based on accelerometry reported an average of 208 minutes/week of MVPA for the EVS measure, whereas inactive participants reported an average of 81 minutes/week (Table 1).

A low positive correlation was found between continuous measures of MVPA reported via the EVS and collected through accelerometry (ρ=0.38, p<0.0001). When restricted to the subsample of 223 participants with <150 minutes/week through accelerometry, the correlation remained positive but to a lower degree (ρ=0.21, p=0.0013). Fair agreement was observed between categories of MVPA based on EVS and accelerometry data (weighted κ=0.29) (Table 2). A slightly higher concordance was found between accelerometry- and EVS-based measures of MVPA when the activity levels were examined dichotomously as <150 or ≥150 minutes/week (simple κ=0.34) (Table 3). When examining the performance of the EVS in identifying inactive participants compared with accelerometry, sensitivity was 67% (95% CI=61%, 73%), specificity was 68% (95% CI=62%, 73%), positive predictive value was 61% (95% CI=55%, 67%), and NPV was 73% (95% CI=68%, 78%) (Table 3).

Table 2.

Frequency and Percent Agreement of Accelerometry Categories of Physical Activity by EVS Categories of Physical Activity (Weighted Kappa=0.29)

Physical activity categorization by EVSb Physical activity categorization by accelerometrya
Inactive
n (%)		 Low active
n (%)		 Moderately active
n (%)		 Active
n (%)		 Overall
n
Inactive 21 (38.9) 12 (28.6) 25 (19.7) 24 (8.1) 82
Low active 6 (11.1) 2 (4.8) 10 (7.9) 10 (3.4) 28
Moderately active 16 (29.6) 16 (38.1) 41 (32.3) 62 (20.8) 135
Active 11 (20.4) 12 (28.6) 51 (40.2) 202 (67.8) 276
Overall, n 54 42 127 298 521
Open in a new tab
a

For accelerometry, inactive defined as 0–29 minutes/week; low active defined as 30–59 minutes/week; moderately active defined as 60–149 minutes/week; active defined as ≥150 minutes/week.

b

For EVS, inactive defined as 0–9 minutes/week; low active defined as 10–59 minutes/week; moderately active defined as 60–149 minutes/week; active defined as ≥150 minutes/week.

EVS, exercise vital sign.

Table 3.

Frequency and Validity of Dichotomized Accelerometry and EVS Categories of Physical Activity (Simple Kappa=0.34)

Physical activity categorization by EVSa Physical activity categorization by accelerometrya
Inactive
N (%)		 Active
N (%)		 Total
N
Inactive 149 (66.8) 96 (32.2) 245
Active 74 (33.2) 202 (67.8) 276
Total, n 223 298 521
Open in a new tab

Notes: Sensitivity: 67% (95% CI=61%, 73%); Specificity: 68% (95% CI=62%, 73%); Positive predictive value (PPV): 61% (95% CI=55%, 67%); Negative predictive value (NPV): 73% (95% CI=68%, 78%).

a

Inactive defined as 0–149 minutes per week. Active defined as ≥150 minutes per week.

EVS, exercise vital sign.

DISCUSSION

When examining the criterion validity of the EVS against 7-day accelerometry in a large cohort of adults, a low-to-fair correlation for weekly MVPA was found, which is consistent with comparisons of previous studies of the EVS and other self-report instruments with accelerometry.15,16,26-28 The EVS correctly classified 68% and 67% of adults who did and did not meet PA guidelines, respectively. However, there was lower agreement between the EVS and accelerometry when further classifying levels of activity. To improve the performance of the EVS, it may be necessary to modify the instrument or supply providers with additional questions to accurately identify adults in lower activity level groups.

Numerous instruments are available for PA self-report29-31; however, the selection of an appropriate instrument requires the consideration of the purpose of the assessment, the context in which it is being conducted, the types of PA behavior that are of interest, and which population groups are the respondents.32-35 It is also necessary to understand how the instrument will be administered, scored, analyzed, and interpreted.32 To assess PA in a clinical setting for the purpose of initiating provider counseling, the optimal instrument would be brief, contain constructs understood by individuals with varied demographic profiles and disease conditions, be quickly scored, and provide easily interpretable results. The EVS meets these criteria and, through this study, has been shown to have criterion validity comparable to other assessment tools used in primary care.13,27

This study particularly focused on assessing the validity of the EVS for the identification of individuals with low activity levels, as these individuals are not meeting recommended activity levels and may benefit from provider counseling. Overall, the EVS correctly classified 67% of participants who were inactive; however, the agreement was lower in correctly classifying adults based on their level of insufficient activity (i.e., as inactive, low active, and moderately active). In addition, a significant number of participants did not meet guidelines according to the accelerometry data were classified as active according to EVS measurement. These individuals likely do not receive counseling and represent an unmet need. Further research is needed to determine whether additional EVS questions or follow-up questions for healthcare providers are needed to more accurately ascertain MVPA status, especially among those who report <150 minutes/week.

Limitations

When interpreting these results, it is also important to acknowledge that objective and self-report data may assess different constructs. Though accelerometers capture minutes of MVPA whether purposeful or not, the EVS specifically asks about days/week and minutes/day of moderate-to-strenuous exercise. Thus, a brisk walk to cross a street would be captured by the accelerometer; however, people will most likely not consider this as PA when responding to the EVS questions. Although accelerometry is often considered to be the gold standard for measuring PA in free-living settings, this is a commonly accepted limitation of using accelerometry to validate self-report instruments.36,37 Despite this, the correlation observed here is similar to the findings of previously reported validity studies that used more comprehensive self-report PA instruments.38-40

There are additional limitations to note when interpreting these results. First, misclassification bias may have occurred during categorization of the EVS and the accelerometer data. For example, in the present study, a person reporting 30 minutes of weekly MVPA for the EVS whose accelerometry indicated 29 minutes of activity would have been classified into different categories. Second, the EVS measurement date often did not align closely with the accelerometer data collection period. The accelerometry data collection was part of a research study, whereas EVS measurements were collected during clinical visits that occurred over a 2-year span prior to study participation. To examine the impact of this approach, a sensitivity analysis using only the EVS measure closest to the date of accelerometer wear was conducted; results of those analyses were similar to those reported here. Finally, this large cohort of adults had a high prevalence of meeting PA guidelines based on the accelerometer data and the EVS. This may be due to the eligibility criteria for PA monitoring in the research study, which required that participants be capable of PA. Comparisons with the overall KPNW membership showed that the study population had a higher mean age but similar levels of comorbidity; however, it is possible that the participants may have been generally healthier than the underlying population from which they were selected.

CONCLUSIONS

The EVS correctly identified the majority of adults who did and did not meet PA guidelines, despite lower levels of agreement between EVS and accelerometry measurements when further classifying levels of activity. Additional research may identify how the EVS tool could be modified or coupled with additional healthcare provider questions to increase its validity and utility in PA assessment, especially among those not meeting recommended levels of activity.

ACKNOWLEDGMENTS

This study is funded by a grant from the National Institute of Diabetes and Digestive and Kidney Diseases, NIH, to Dr. Fortmann (R01 DK103385).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

No financial disclosures were reported by the authors of this paper.

REFERENCES

  • 1.2018 Physical activity guidelines advisory committee scientific report. Office of Disease Prevention and Health Promotion. https://health.gov/paguidelines/second-edition/report/. Published 2018. Accessed December 4, 2018. [Google Scholar]
  • 2.Nutrition, physical activity, and obesity: data, trend and maps. Centers for Disease Control and Prevention. https://www.cdc.gov/nccdphp/dnpao/data-trends-maps/index.html. Accessed November 20, 2019. [Google Scholar]
  • 3.Krist AH, Davidson KW, Mangione CM, et al. Behavioral counseling interventions to promote a healthy diet and physical activity for cardiovascular disease prevention in adults with cardiovascular risk factors: US Preventive Services Task Force recommendation statement. JAMA. 2020;324(20):2069–2075. 10.1001/jama.2020.21749. [DOI] [PubMed] [Google Scholar]
  • 4.Curry SJ, Krist AH, Owens DK, et al. Behavioral weight loss interventions to prevent obesity-related morbidity and mortality in adults: US Preventive Services Task Force recommendation statement. JAMA. 2018;320(11):1163–1171. 10.1001/jama.2018.13022. [DOI] [PubMed] [Google Scholar]
  • 5.Healthy People 2030. HHS. https://health.gov/healthypeople/objectives-and-data/browse-objectives/overweight-and-obesity/increase-proportion-health-care-visits-adults-obesity-include-counseling-weight-loss-nutrition-or-physical-activity-nws-05. Publsihed 2020. Accessed October 15, 2020. [Google Scholar]
  • 6.Kraus WE, Bittner V, Appel L, et al. The National Physical Activity Plan: a call to action from the American Heart Association: a science advisory from the American Heart Association. Circulation. 2015;131(21):1932–1940. 10.1161/cir.0000000000000203. [DOI] [PubMed] [Google Scholar]
  • 7.Sallis RE, Matuszak JM, Baggish AL, et al. Call to action on making physical activity assessment and prescription a medical standard of care. Curr Sports Med Rep. 2016;15(3):207–214. 10.1249/jsr.0000000000000249. [DOI] [PubMed] [Google Scholar]
  • 8.National Academies of Sciences Engineering and Medicine. Implementing Strategies to Enhance Public Health Surveillance of Physical Activity in the United States. Washington, DC: National Academies Press (US); 2019. 10.17226/25444. [DOI] [PubMed] [Google Scholar]
  • 9.van Sluijs EM, van Poppel MN, Twisk JW, Chin APMJ, Calfas KJ, van Mechelen W. Effect of a tailored physical activity intervention delivered in general practice settings: results of a randomized controlled trial. Am J Public Health. 2005;95(10):1825–1831. 10.2105/ajph.2004.044537. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Goldstein MG, Pinto BM, Marcus BH, et al. Physician-based physical activity counseling for middle-aged and older adults: a randomized trial. Ann Behav Med. 1999;21(1):40–47. 10.1007/bf02895032. [DOI] [PubMed] [Google Scholar]
  • 11.Norris SL, Grothaus LC, Buchner DM, Pratt M. Effectiveness of physician-based assessment and counseling for exercise in a staff model HMO. Prev Med. 2000;30(6):513–523. 10.1006/pmed.2000.0673. [DOI] [PubMed] [Google Scholar]
  • 12.Grandes G, Sanchez A, Montoya I, Ortega Sanchez-Pinilla R, Torcal J. Two-year longitudinal analysis of a cluster randomized trial of physical activity promotion by general practitioners. PloS One. 2011;6(3):e18363. 10.1371/journal.pone.0018363. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Coleman KJ, Ngor E, Reynolds K, et al. Initial validation of an exercise “vital sign” in electronic medical records. Med Sci Sports Exerc. 2012;44(11):2071–2076. 10.1249/mss.0b013e3182630ec1. [DOI] [PubMed] [Google Scholar]
  • 14.Young DR, Coleman KJ, Ngor E, Reynolds K, Sidell M, Sallis RE. Associations between physical activity and cardiometabolic risk factors assessed in a Southern California health care system, 2010–2012. Prev Chronic Dis. 2014;11:E219. 10.5888/pcd11.140196. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Fitzgerald L, Ozemek C, Jarrett H, Kaminsky LA. Accelerometer validation of questionnaires used in clinical settings to assess MVPA. Med Sci Sports Exerc. 2015;47(7):1538–1542. 10.1249/mss.0000000000000565. [DOI] [PubMed] [Google Scholar]
  • 16.Joseph RP, Keller C, Adams MA, Ainsworth BE. Validity of two brief physical activity questionnaires with accelerometers among African-American women. Prim Health Care Res Dev. 2016;17(3):265–276. 10.1017/s1463423615000390. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Frank LD, Kuntz JL, Chapman JE, et al. The health and economic effects of light rail lines: design, methods, and protocol for a natural experiment. BMC Public Health. 2019;19:200. 10.1186/s12889-019-6518-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Leo MC, Lindberg NM, Vesco KK, Stevens VJ. Validity of medical chart weights and heights for obese pregnant women. EGEMS (Wash DC). 2014;2(1):1051. 10.13063/2327-9214.1051. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Smith N, Coleman KJ, Lawrence JM, et al. Body weight and height data in electronic medical records of children. Int J Pediatr Obes. 2010;5(3):237–242. 10.3109/17477160903268308. [DOI] [PubMed] [Google Scholar]
  • 20.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373–383. 10.1016/0021-9681(87)90171-8. [DOI] [PubMed] [Google Scholar]
  • 21.Jarrett H, Fitzgerald L, Routen AC. Interinstrument reliability of the ActiGraph GT3X+ ambulatory activity monitor during free-living conditions in adults. J Phys Act Health. 2015;12(3):382–387. 10.1123/jpah.2013-0070. [DOI] [PubMed] [Google Scholar]
  • 22.Santos-Lozano A, Marin PJ, Torres-Luque G, Ruiz JR, Lucia A, Garatachea N. Technical variability of the GT3X accelerometer. Med Eng Phys. 2012;34(6):787–790. 10.1016/j.medengphy.2012.02.005. [DOI] [PubMed] [Google Scholar]
  • 23.John D, Freedson P. ActiGraph and Actical physical activity monitors: a peek under the hood. Med Sci Sports Exerc. 2012;44(1S):S86–S89. 10.1249/mss.0b013e3182399f5e. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, McDowell M. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc. 2008;40(1):181–188. 10.1249/mss.0b013e31815a51b3. [DOI] [PubMed] [Google Scholar]
  • 25.Hastings C Jr, Mosteller F, Tukey J, Winsor C. Low moments for small samples: a comparative study of order statistics. Annals of Mathematical Statistics. 1947;18(3):413–426. 10.1214/aoms/1177730388. [DOI] [Google Scholar]
  • 26.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. 10.1186/1471-2458-14-1255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Smith TO, McKenna MC, Salter C, et al. A systematic review of the physical activity assessment tools used in primary care. Fam Pract. 2017;34(4):384–391. 10.1093/fampra/cmx011. [DOI] [PubMed] [Google Scholar]
  • 28.Meriwether RA, McMahon PM, Islam N, Steinmann WC. Physical activity assessment: validation of a clinical assessment tool. Am J Prev Med. 2006;31(6):484–491. 10.1016/j.amepre.2006.08.021. [DOI] [PubMed] [Google Scholar]
  • 29.Pereira MA, FitzerGerald SJ, Gregg EW, et al. A collection of Physical Activity Questionnaires for health-related research. Med Sci Sports Exerc. 1997;29(6 suppl):S1–S205. [PubMed] [Google Scholar]
  • 30.Physical Activity Surveillance Systems. Centers for Disease Control and Prevention. https://www.cdc.gov/physicalactivity/data/surveillance.htm#Assessing. Published 2017. Accessed December 9, 2019. [Google Scholar]
  • 31.Physical Activity Self-Report Measures Descriptions 2020. http://www.parcph.org/subjDefList.aspx. Accessed June 1, 2020.
  • 32.Ainsworth BE, Caspersen CJ, Matthews CE, Masse LC, Baranowski T, Zhu W. Recommendations to improve the accuracy of estimates of physical activity derived from self report. J Phys Act Health. 2012;9(suppl 1):S76–S84. 10.1123/jpah.9.s1.s76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Sternfeld B, Goldman-Rosas L. A systematic approach to selecting an appropriate measure of self-reported physical activity or sedentary behavior. J Phys Act Health. 2012;9(suppl 1):S19–S28. 10.1123/jpah.9.s1.s19. [DOI] [PubMed] [Google Scholar]
  • 34.Golightly YM, Allen KD, Ambrose KR, et al. Physical activity as a vital sign: a systematic review. Prev Chronic Dis. 2017;14:E123. 10.5888/pcd14.170030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Wald A, Garber CE. A review of current literature on vital sign assessment of physical activity in primary care. J Nurs Scholarsh. 2018;50(1):65–73. 10.1111/jnu.12351. [DOI] [PubMed] [Google Scholar]
  • 36.Haskell WL. Physical activity by self-report: a brief history and future issues. J Phys Act Health. 2012;9(suppl 1):S5–S10. 10.1123/jpah.9.s1.s5. [DOI] [PubMed] [Google Scholar]
  • 37.Troiano RP, McClain JJ, Brychta RJ, Chen KY. Evolution of accelerometer methods for physical activity research. Br J Sports Med. 2014;48(13):1019–1023. 10.1136/bjsports-2014-093546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Wanner M, Probst-Hensch N, Kriemler S, Meier F, Autenrieth C, Martin BW. Validation of the long international physical activity questionnaire: influence of age and language region. Prev Med Rep. 2016;3:250–256. 10.1016/j.pmedr.2016.03.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Sternfeld B, Gabriel KP, Jiang SF, et al. Risk estimates for diabetes and hypertension with different physical activity methods. Med Sci Sports Exerc 2019;51(12):2498–2505. 10.1249/mss.0000000000002083. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.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. 10.1249/mss.0b013e3182a0595f. [DOI] [PubMed] [Google Scholar]

RESOURCES