Objective Measures of Physical Activity in Rural Communities: Factors Associated With a Valid Wear and Lessons Learned

Click name to view affiliation

Amanda Gilbert
Search for other papers by Amanda Gilbert in
Current site
Google Scholar
PubMed
Close
,
Alan Beck
Search for other papers by Alan Beck in
Current site
Google Scholar
PubMed
Close
,
Natalicio Serrano
Search for other papers by Natalicio Serrano in
Current site
Google Scholar
PubMed
Close
, and
Ross C. Brownson
Search for other papers by Ross C. Brownson in
Current site
Google Scholar
PubMed
Close
Open access

Background: Compared with urban/suburban counterparts, rural communities experience lower rates of physical activity (PA) and higher rates of chronic disease. Promoting PA is important for disease prevention but requires reliable and valid measurement of PA. However, little is known about effectively collecting objective PA data in rural communities. Using data from a cluster randomized trial (Heartland Moves), which aims to increase PA in rural Missouri, this study explored factors associated with successful objective PA data collection and presents lessons learned. Methods: Baseline survey and accelerometry data were collected through Heartland Moves (n = 368) from August 2019 to February 2021, in southeast Missouri. Chi-square and logistic regression analyses were used to explore factors (demographics, subjective PA, and SMS reminders) associated with valid wear of PA devices. Results: Overall, 77% had valid wears. Participants who were not married (odds ratio [OR] 0.48, 95% confidence interval [CI], 0.30–0.79) and those living alone (OR 0.49, 95% CI, 0.30–0.81) were less likely to have valid wears. Participants who met PA guidelines (OR 1.69, 95% CI, 1.03–2.75) or received SMS reminders (OR 3.25; 95% CI, 1.97–5.38) were more likely to have valid wears. Conclusions: Results are supported by lessons learned, including importance of communication (SMS reminders), accessing hard-to-reach groups (living alone), and need to adapt during data collection.

A significant body of research has found physical activity (PA) to be beneficial in the prevention and treatment of myriad chronic diseases, as well as for overall well-being and mental health.13 The relationship between PA and health is especially important for rural communities, where micropolitan rural areas have the lowest rate of PA among any other subgroup,4 and rates of chronic disease are higher than in urban and suburban populations.48 The development and testing of interventions to promote PA in rural communities for chronic disease prevention requires reliable and valid measurement of PA; however, little is known about effectively collecting objective PA data in rural communities.

The PA measurement is important for understanding the relationship between PA and health, testing effective interventions and policies for promoting PA, and identifying which populations are most in need of these PA promoting policies. PA measurement provides an understanding of what types (ie, aerobic, muscle strengthening); domains (ie, leisure time, travel); and dose (ie, frequency, duration, intensity) of PA will result in improved health outcomes for various populations.1 These PA measures inform recommendations for beneficial levels of PA (eg, 2018 Physical Activity Guidelines), which have substantial implications for PA research, policy, and practice.1,912 Measuring PA outcomes or effectiveness of PA interventions also guides PA policy development.13 PA measures are also used for surveillance and monitoring to provide valuable information about which populations have low levels of PA and may benefit most from policies promoting PA.1,9

Researchers generally measure PA in 2 ways—subjectively, through self-report (surveys, in-person interviews, and telephone interviews) and objectively, with devices worn by participants. Self-report can provide information about PA type and domain, as well as barriers and facilitators to being active. Self-report also has utility as a source for large-scale PA surveillance (eg, Behavioral Risk Factor Surveillance Survey, National Health Information Survey). Although a valuable source of PA data, subjective measures are often prone to issues with recall bias and social desirability bias, systematically leading to overestimation of PA dose.14,15 This is of particular concern, since dose (frequency, duration, and intensity) is strongly associated with health outcomes.1,16 Objective measures of PA, which record PA in real time through devices worn on the body, provide more accurate data on frequency and time spent being physically active.17,18 These devices can also be paired with global positioning system (GPS) devices to provide information about where PA is happening. While self-report is useful and should be a part of PA measurement, objective measurement through devices can offer a more reliable and valid measure of PA, as it might relate to health outcomes.

Although vital to PA measurement, obtaining objective measures of PA can be challenging. Studies have explored barriers and facilitators to collecting objective PA data from devices including the role of health behaviors,19,20 sociodemographic characteristics (eg, age, race, gender, ethnicity),2022 and body mass index. While this research provides insight into the challenges for objective PA data collection, little is known about the unique barriers and facilitators to collecting objective PA data from devices in rural communities, and more specifically, data collection for intervention studies conducted in rural populations.

Understanding objective PA data collection is an important gap to address since rural communities may face challenges associated with geographic distance and building community trust.23 Geographic distances can limit in-person contact, requiring the mailing of devices and reliance on phone and text message for communication. Although some large-scale surveillance studies also mail devices, the reliance on phone and text message communication in the context of a rural intervention study may make it more difficult to overcome community mistrust of researchers. This distance may also make adapting to unforeseen or unknown barriers more difficult and connecting with harder to reach subpopulations (ie, older, less educated) much more challenging than for in-person data collection. Challenges around objective PA data collecting in rural communities not only has implications for promoting PA but also can make testing PA interventions less efficient. Mailing out devices is costly and time consuming, making it critical each device mailed to a participant has the highest likelihood of coming back with valid wear time.24

Therefore, the aims of this study were to use data from a group randomized controlled trial (Heartland Moves), which aims to increase PA in rural southeast Missouri,25 to explore factors associated with successful objective PA data collection and to compliment these results with a presentation of lessons learned during the data collection process.

Methods

Participants and Procedures

This study represents the baseline findings from Heartland Moves, a large PA intervention aimed at increasing PA among rural participants.25 Adult participants (ie, 18 years of age or older) were recruited from 14 rural communities in southeastern Missouri via address-based sampling, word-of-mouth, and at in-person community events. Rurality was operationally defined as a Rural-Urban Continuum Code (RUCC) of 4 or greater.26 It is important to note, one community added on later due to lack of survey response had a RUCC of 3; however, it was paired with a community of a similar demographic makeup (ie, population size, poverty rate, minority population), and a RUCC of 4. Participants who consented to being in the Heartland Moves study completed a telephone survey to collect baseline data. At the end of the survey, participants were recruited to wear accelerometer (ActiGraph, GT3x+, Pensacola, FL) and GPS devices (QSTARZ BT-Q1000XT) for 1 week. Research staff called all participants who reported interest in wearing devices to explain the procedures and obtain consent for device wear. If the participant was still interested and gave consent, they were sent the devices via the US Postal Service with instructions. Participants were also asked if they would like to receive a short message service (ie, text message, SMS) reminders during their week of wear. Upon completion of wearing the devices, participants were offered a $25 incentive. On average, the cost of mailing devices to one participant was $60, with each time devices are sent out lasting on average between 15 and 20 days.

A total of 1241 participants were surveyed in the catchment area, with 718 participants expressing interest in wearing the devices. When contacted by research staff, 181 could not be reached, 5 were not eligible (due to not living in the area any longer), 118 declined to participate, 414 agreed to participate; however, of those who agreed to participate, 44 people sent the devices back without wearing them, and 2 devices were lost in the mail. See Figure 1 for a flow diagram of participants. The institutional review board at Washington University in St. Louis (IRB number 201809089) approved all procedures.

Figure 1
Figure 1

—Flow diagram of accelerometry data collection.

Citation: Journal of Physical Activity and Health 19, 4; 10.1123/jpah.2021-0677

Measures

Individual Factors

Data on individual factors were obtained through a telephone survey administered during baseline data collection of the Heartland Moves study. Given the exploratory nature or this analysis, individual factors were chosen based on what was available through the telephone survey, could potentially be associated with valid wears, or was previously assessed in studies focused on other populations.1922,24,2729 Participants were asked about their age (18–34 y, 35–65 y, and more than 65 y); race (White, Black or American Indian Alaska Native, or multiracial); gender (male and female); marital status (married or member of an unmarried couple, divorced, widowed, separated, or never married); living situation (with relatives or nonrelatives, alone); educational attainment (less than high school or high school, some college or Associates degree, Bachelor’s degree or higher); annual household income (less than or equal to $35,000, $35,001–$75,000, $75,001–$100,000, more than $100,000); and employment type (employed and not employed).

The PA was measured in 2 ways. First, participants were asked to report whether they currently exercised (yes and no) utilizing the RM 1-FM Physical Activity Stages of Change–Questionnaire.30 For this variable, the term “physical activity” was replaced with “exercise” in response to cognitive response testing conducted in our study population. Second, weekly minutes of moderate and vigorous PA were calculated using the Global Physical Activity Questionnaire.31 The Global Physical Activity Questionnaire asks about PA dose (ie, frequency, duration, intensity) as well as domain (ie, work, recreational, commute) for a more comprehensive measure of PA. A dichotomous variable was then created based on whether participants met aerobic PA guidelines (ie, 150 min of moderate intensity and/or 75 min of vigorous activity per week).32

The SMS measure was dichotomous (did or did not receive SMS reminders during their week of wear). Receiving SMS reminders to wear the accelerometer and GPS devices was optional for participants. When participants were contacted to go over instructions for wearing devices, they were offered the option to receive daily SMS reminders to wear the devices and charge the GPS device. For example, one reminder read “Good morning! Please remember to wear the device belt with both devices all day until bedtime and charge the black GPS device at night.”

The COVID-19 pandemic occurred during the end of baseline data collection. Therefore, a measure on whether data collection occurred prepandemic or during the pandemic was included to explore whether the COVID-19 pandemic had an effect on participants obtaining valid wear times. Participants who wore devices prior to March 11, 2020 were considered to have wear time prior to the COVID-19 pandemic, while participants who wore devices after March 11, 2020 were considered to have wear time during the COVID-19 pandemic.

Valid Wear Time

Successful objective data collection for each participant was based on a threshold for valid wear time, which was informed by best practices for objective PA data collection.33 Both accelerometer and GPS devices were worn together on a belt. Accelerometer and GPS data were collected for most of baseline data collection with a daily threshold of 10 hours. However, due to the low number of valid wears, the research team made the decision to reduce the minimum wear to 8 hours per day based upon the data collected (eg, common wear time, sample characteristics) to that point in time, and other literature utilizing an 8-hour threshold.34,35 Accelerometer data were downloaded with 1-second epoch levels into ActiLife software (Actigraph Corporation, Pensacola, FL) using Troiano 2007 parameters.15 Nonwear time was defined as a run of zeros for 60 minutes. Participants were considered to have a valid wear time if the time worn met the threshold criteria of 8 hours per day, for at least 3 days. Valid wear time was measured as a dichotomous variable (yes, the time worn met the threshold criteria and no, the time worn did not meet the threshold criteria).

Statistical Analysis

Frequencies and percentages were calculated for valid wear, and all individual and COVID-19 factors. Chi-square tests and logistic regressions were used to explore associations between individual factors (age, race, gender, marital status, living situation, education, income, employment, exercise, meeting PA guidelines, and SMS reminders), and data collection during COVID-19 with valid wear time. Assumptions of sample size were met for chi-square analyses and of multicollinearity for logistic regressions. All data analysis was conducted using SPSS (version 25; IBM Corp Armonk, NY).36

Results

Sample Characteristics

The analytic sample consisted of 368 participants who wore the accelerometer and GPS devices. (Table 1). Of those participants, 284 (77%) had a valid wear time and 84 (23%) did not. Among the 284 who attained a valid wear, 23% objectively met PA guidelines. Around half of participants were between the ages of 35 and 65 years old (57%), married or part of an unmarried couple (59%), and employed (52%). Most were White (87%), female (73%), and living with relatives or nonrelatives (67%). Around 24% of participants had an educational attainment of a high school diploma or less and 25% of participants had an annual household income less than $35,000. Most participants reported exercising (69%), and based on a self-report measure of PA (Global Physical Activity Questionnaire), 60% met PA guidelines. In terms of communication, 67% of participants opted to receive SMS reminders. Considering the impact of the COVID-19 pandemic on data collection, most of the participants wore devices prior to the pandemic (82%).

Table 1

Frequencies of Individual and COVID-19 Factors by Accelerometer Valid Wear Time Among Rural Participants in the Heartland Moves Study

Total (N = 368)aValid wear time (n = 284, 77.2%)bNonvalid wear time (n = 84, 22.8%)b
Individual factors   
 Age, n (%)   
  18–34 y51 (13.8)38 (74.5)13 (25.5)
  35–65 y209 (56.8)166 (79.4)42 (20.1)
  Greater than 65 y109 (29.6)80 (73.4)29 (26.6)
 Race, n (%)   
  White319 (86.9)248 (78.0)71 (22.3)
  Black, American Indian, and Alaska Native, multiracial49 (13.1)36 (73.5)13 (26.5)
 Gender, n (%)   
  Female267 (72.6)206 (77.2)61 (22.8)
  Male101 (27.4)78 (77.2)23 (22.8)
 Marital status, n (%)   
  Married, member of an unmarried couple217 (58.7)179 (82.5)38 (17.5)
  Divorced, widowed, separated, never married153 (41.3)105 (64.7)46 (35.3)
 Living situation, n (%)   
  With relatives or nonrelatives248 (67.4)202 (81.5)46 (18.5)
  Alone120 (32.6)82 (68.3)38 (31.7)
 Education, n (%)   
  Less than high school, high school/GED88 (23.9)64 (72.7)24 (27.2)
  Some college/vocational, associates degree129 (35.1)95 (73.6)34 (26.4)
  Bachelor’s degree or higher151 (41.0)125 (82.8)26 (17.2)
 Income, n (%)   
  Less than or equal to $35,000133 (24.7)95 (71.4)38 (28.6)
  $35,001–$75,000112 (33.6)90 (80.4)22 (19.6)
  $75,001–$100,00049 (14.7)40 (81.6)9 (18.4)
  Greater than $100,00037 (11.1)30 (81.1)7 (18.9)
 Employment, n (%)   
  Employed189 (51.5)146 (77.2)43 (22.8)
  Not employed178 (48.5)138 (77.5)40 (22.5)
 Currently exercise, n (%)   
  Yes254 (68.6)196 (77.8)56 (22.2)
  No116 (31.4)88 (75.9)28 (24.1)
 Meets physical activity guidelines,c n (%)   
  Meets guidelines220 (59.8)177 (80.8)42 (19.2)
  Does not meet guidelines148 (40.2)105 (71.4)42 (28.6)
 SMS reminders   
  Received SMS245 (66.6)207 (84.5)38 (15.5)
  Did not receive SMS123 (33.4)77 (62.6)46 (37.4)
COVID-19   
 Before or during COVID-19, n (%)   
  Before COVID-19305 (82.4)230 (75.9)73 (24.1)
  During COVID-1965 (17.6)54 (83.1)11 (16.9)

aPercentages in this column are based on entire sample (column). bPercentages in this column are based on category (row). cMeasured using the Global Physical Activity Questionnaire.

Correlates

We found significant associations for the demographic predictors of marital status, χ2 = 8.47; P = .004, living situation, χ2 = 7.90; P = .005, and meeting PA guidelines, χ2 = 4.39; P = .04. Compared with participants who were married or part of an unmarried couple, participants who were divorced, widowed, separated, or never married had 48% lesser odds of having a valid wear time (odds ratio [OR] = 0.48; 95% confidence interval [CI], 0.30−0.79) (Table 2). Regarding living situation, participants who lived alone had 49% lesser odds of having a valid wear time than participants who lived with relatives or nonrelatives (OR 0.49; 95% CI, 0.30−0.81). For PA, participants who met PA guidelines had 1.69 times greater odds of having a valid wear time than participants who did not meet PA guidelines (OR 1.69; 95% CI, 1.03−2.75). We also found significant associations with the study design predictor of receiving SMS reminders, χ2 = 22.27; P < .001. Participants who received SMS reminders had 3.25 times greater odds of having a valid wear time, than participants who did not receive SMS reminders (OR = 3.25; 95% CI, 1.97−5.38).

Table 2

Analysis of Individual and COVID-19 Factors With Accelerometer Valid Wear Time Among Rural Participants in the Heartland Moves Study (N = 368)

Valid weara
ORb95% CI
Individual factors  
 Age, y  
  Reference: 18–34
  35–651.350.66–2.76
  Greater than 650.940.44–2.02
 Race  
  Reference: White
  Black, American Indian and Alaska Native, multiracial0.820.41–1.67
 Gender  
  Reference: male
  Female1.000.58–1.72
 Marital statusc  
  Reference: married, member of unmarried couple
  Divorced, widowed, separated, never married0.48**0.30–0.79
 Living situationc  
  Reference: with relatives or nonrelatives
  Alone0.49**0.30–0.81
 Education  
  Reference: less than high school, high school/GED
  Some college/vocational, associates degree1.050.57–1.93
  Bachelor’s degree or higher1.800.96–3.39
 Income  
  Reference: less than or equal to $35,000
  $35,001–$75,0001.640.90–3.00
  $75,001–$100,0001.780.79–4.02
  Greater than $100,0001.710.69–4.24
 Employment  
  Reference: employed
  Not employed1.020.62–1.66
 Currently exercise  
  Reference: no
  Yes1.110.66–1.87
 Meets PA guidelinesc  
  Reference: no
  Yes1.69*1.03–2.75
 SMS remindersc  
  Reference: no
  Yes3.25**1.97–5.38
COVID-19 factor  
 Reference: before or during COVID-19
 During COVID-191.560.77–3.14

Abbreviations: CI, confidence interval; OR, odds ratio; PA, physical activity.

aValid wear—time worn met the threshold criteria of 8 hours per day for at least 3 days. bORs are unadjusted. cSignificant association from chi-squared analyses (P < .05).

Statistical Significance at the *P < .05, **P < .01 level.

Discussion

This study explored factors associated with successful objective PA data collection in rural communities and lessons learned during this process. At the end of baseline data collection, 77% of participants who wore devices had a valid wear time. Given wear time criteria varies by study, it is difficult to interpret whether 77% is higher, lower, or average for valid wears obtained through objective PA data collection. Using similar wear time criteria, Evenson et al27 assessed accelerometer data collection in a cohort of US Hispanic adults and obtained around the same percentage of valid wears (78%). Although not in a rural community, this study also collected objective data in a hard-to-reach group, supporting the 78% of valid wears as a strong comparison for our study. Other studies using accelerometers in adult populations had similar rates of valid wears for 4 or more days of wear time, ranging from 70% to 95%.19,21,22,24

Some of our findings around correlates of valid wear time are in line with these previous studies. Similar to our results for married participants or those living with others, previous studies found participants who were married or partnered were more likely to have a valid wear time.20,27 In regard to health behaviors, a study conducted by Loprinzi et al20 also found participants with lower levels of self-reported PA were less likely to have valid wear times. In contrast to our findings, other studies have found differences in valid wear by race/ethnicity,20 age,20,22,28,29 gender,27 income,27 and education.20 These differences in our findings regarding demographic correlates for valid wear may be due to the populations sampled, sample size, or variations in valid wear criteria. It may also reflect potential differences for data collection within some rural populations. In this sample, most participants were white, which may also influence the results. More research will be needed to explore the relationship between valid wears and sociodemographic factors and whether they are meaningful correlates in rural settings.

Based on our findings, we present 3 important considerations—importance of communication, strategies for hard-to-reach groups, and importance of adapting during data collection for conducting data collection in rural communities. These areas of importance are informed both by the data analyzed in this paper and experiences of the research team during data collection.

Importance of Communication

One of the most important considerations during data collection was the importance of communication. Best practices for objective PA data collection highlight the importance of communication, but we found this is likely to be even more important in rural communities, when geographic distances often necessitate data collection through mailing of devices, limiting communication to phone calls and text messages. With the inability to meet for in-person instruction on device wear, it is imperative to maintain contact to promptly address questions about device wear and to keep participants engaged and motivated. In this study, we found participants who elected to have SMS reminders were 3 times more likely to attain the minimum wear time. While the typical recommendation for communication is telephone calls, when there are over 100 devices in the field, it can be difficult to make frequent calls.37 There are myriad benefits of using text messaging as reminders: (1) ease of use for the sender, messages can be sent to multiple participants quickly; (2) ease of use by the addressee, the message can be read by the addressee on their own time; and (3) addressing concerns, a call can be arranged if issues arise. There may still be people without cellular phones in which text messaging would not be feasible; however, cell phones are becoming ubiquitous among urban and rural residents alike.38 One potential way to increase participants receiving text message reminders is to make receiving text messages the default. Based on research from behavioral economics and nudge theory, there is strong evidence that people are more likely to stick with the default option rather than opt out.39 This nudge and explaining the importance of text message reminders to participants and reason for the default option may increase acceptance of text message reminders.40 Given our findings, use of text messages for communication can improve the chances of collecting valid data in rural communities where in-person contact is limited.

Strategies for Hard-to-Reach Groups

Another important consideration is finding strategies for successful data collection within hard-to-reach groups (eg, living alone, less physically active). This consideration comes from a few important observations from our study about potentially harder to reach groups within rural communities. First, participants who lived with a support person/group (eg, married, lives with others) were more likely to have a valid wear time than those who lived alone or did not have a partner. Roughly 30% of our participants lived alone and around 40% were not married or living with a partner. Through communication with participants, we learned those who often spoke to their friends/family about the devices and their participation in the study, tended to feel more involved, which perhaps held them more accountable to wearing the devices. Given these findings, we suggest a strategy of engaging participant social support for device wear. For example, when sending out devices, determine if the potential participant has a support group or person to help with adherence to the protocol. Second, we found participants meeting PA guidelines via subjective measures were more likely to attain a valid wear of the devices. It might also be the case that participants who are more active, may be more invested or interested in the PA device data and thus more likely to wear the devices. Our research team did not plan to send out feedback on PA to each participant; however, many participants reached out to determine how well they did. We would recommend this as a strategy to others to be prepared to receive inquiries regarding participant PA and have a standard practice by which to report PA back (eg, comparing output to PA guidelines). Reporting PA following device wear may motivate participants to meet the wear time threshold for a valid wear, especially those who are less physically active but should be balanced with the potential bias introduced through the Hawthorne effect.

Adaptation During Data Collection

Finally, being able to adapt during data collection is key, especially since little research has been done around best practices for data collection in rural communities. This was particularly important for determining the appropriate time required for a valid wear, which has not formally been tested within rural communities, though studies in other vulnerable populations (ie, Latinas) have used similar wear time when also collecting objective data.41,42 During data collection, a low number of valid wear’s necessitated a reevaluation of our valid wear measurement. While our research team was hopeful to get a minimum of 10 hours of valid wear, the participants were wearing both an accelerometer and GPS device; therefore, we followed common practice to use lower thresholds to attain a valid wear.41,42 Participants were still required to attain at least 3 valid days per week. We first adjusted the nonwear time from 60-minute run of zeros to 120 minutes, which did not increase wear time. Our research team then decided to decrease the minimum wear time from 10 hours per day to 8 hours per day, which resulted in increased valid wears. We learned rural adults may be less likely to wear the device for extended periods during the day; therefore, it may be prudent to reduce the wear time for rural adults. We also learned that many older adults are extremely sedentary, whereby they may have worn the devices for well over 12 hours; however, if they spend most of their time not moving (eg, sitting in a chair for an extended period, napping) the reduced wear time requirement seemed to help attain a valid wear.

We had to adapt data collection during the COVID-19 pandemic. While not a consideration unique to rural communities, we relied heavily on the research team’s local community health coordinator and community collaborators to assess local response and experience of COVID-19, and when and how best to continue data collection. Most of baseline data was collected prior to the spread of COVID-19 and associated restrictions in the participating rural communities. While we did not find this impacted valid wear times in this study, we temporarily halted data collection, resent devices which had been in the field when restrictions took place, and during initial phone calls for device instruction, created an open dialoge to assess participant concern and comfort level with device wear during the pandemic.

Limitations

While our study offers valuable insights into characteristics among rural communities associated with valid accelerometry wear time, there are a few limitations worth noting. First, we were only able to assess characteristics captured within the baseline survey for the Heartland Moves project. As such, we were unable to explore other factors, which may be pertinent to participants in rural communities successfully wearing accelerometer devices. For example, we did not look at information like occupation. Second, we asked participants to wear both an accelerometer and GPS device during data collection and were unable to assess if or how the addition of the GPS device impacted valid wears. Third, while we were able to explore objective PA data collection in rural communities, it is important to note that the rural population is not homogenous. As such, our findings may be limited in their generalizability to all rural populations in the United States. In addition, one of the communities in the sample had a RUCC of 3, which does not meet the operationalized definition of rurality. However, this community was paired with a community of similar demographic make-up and RUCC of 4. Fourth, we did not explore correlates of whether participants agreed to wear devices and if there were significant differences between these 2 groups of participants. Due to this, our findings are limited to understanding successful objective PA data collection among participants who agree to wear devices. Despite these limitations, our study does offer unique insight into factors associated with valid wear time in rural communities. This is particularly important given lower levels of PA in rural communities and thus a need for more research into ways to promote PA. Furthermore, our study analyzed accelerometry and demographic data from many participants and across multiple rural communities, lending support to generalizability of our findings.

Conclusions

Our study explored characteristics among rural participants which were predictive of a valid accelerometry wear time and presented lessons learned during the data collection process. Our findings highlight 3 important areas of consideration (importance of communication, strategies for hard-to-reach groups, and importance of adapting during data collection) for data collection in rural communities. These considerations are a foundation for future research to measure PA levels and assess PA intervention for rural populations more effectively. Given the low levels of PA in rural communities and high rates of chronic disease associated with inactivity,47 this is imperative. Future studies should continue to explore factors relevant for successful objective PA data collection in underserved populations such as in rural communities. While we have identified some important considerations and strategies for successful data collection, more needs to be understood about ways researchers can best facilitate valid accelerometry wear.

Acknowledgments

The authors would like to thank Linda Dix and Mary Adams for their administrative support and Dixie Duncan for assistance in data collection. Funding Source/Trial Registration: This work was supported by a grant (number R01CA211323) from the National Institutes of Health and National Cancer Institute. In addition, authors were supported by the National Heart, Lung, and Blood Institute training grant (number T32HL130357), the National Cancer Institute training grant (number T32CA057699, the Centers for Disease Control and Prevention (number U48DP006395), and the Foundation for Barnes-Jewish Hospital. The findings and conclusions in this paper are those of the authors and do not necessarily represent the official positions of the National Institutes of Health or the Centers for Disease Control and Prevention. This study is registered at www.clinicaltrials.gov (number NCT03683173).

References

  • 1.

    US Department of Health and Human Services. 2018 Physical Activity Guidelines Advisory Committee. Washington, DC: US Department of Health and Human Services; 2018. doi:10.1111/j.1753-4887.2008.00136.x

    • Search Google Scholar
    • Export Citation
  • 2.

    Moore SC, Lee IM, Weiderpass E, et al. Association of leisure-time physical activity with risk of 26 types of cancer in 1.44 million adults. JAMA Intern Med. 2016;176(6):816825. PubMed ID: 27183032 doi:10.1001/jamainternmed.2016.1548

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3.

    Bize R, Johnson JA, Plotnikoff RC. Physical activity level and health-related quality of life in the general adult population: a systematic review. Prev Med. 2007;45(6):401415. doi:10.1016/j.ypmed.2007.07.017

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4.

    Fan JX, Wen M, Kowaleski-Jones L. Rural–urban differences in objective and subjective measures of physical activity: findings from the national health and nutrition examination survey (NHANES) 2003–2006. Prev Chronic Dis. 2014;11:111. doi:10.5888/pcd11.140189

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5.

    Whitfield GP, Carlson SA, Ussery EN, Fulton JE, Galuska DA, Petersen R. Trends in meeting physical activity guidelines among urban and rural dwelling adults—United States, 2008–2017. MMWR Morb Mortal Wkly Rep. 2019;68(23):513518. PubMed ID: 31194722 doi:10.15585/mmwr.mm6823a1

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6.

    Cohen SA, Greaney ML, Sabik NJ. Assessment of dietary patterns, physical activity and obesity from a national survey: rural-urban health disparities in older adults. PLoS One. 2018;13(12):e0208268. doi:10.1371/journal.pone.0208268

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7.

    Singh GK, Siahpush M. Widening rural-urban disparities in all-cause mortality and mortality from major causes of death in the USA, 1969–2009. J Urban Heal. 2014;91:272292. doi:10.1007/s11524-013-9847-2

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8.

    Bolin JN, Bellamy GR, Ferdinand AO, et al. Rural healthy people 2020: new decade, same challenges. J Rural Heal. 2015;31(3):326333. doi:10.1111/jrh.12116

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9.

    Hallal PC, Andersen LB, Bull FC, et al. Global physical activity levels: surveillance progress, pitfalls, and prospects. Lancet. 2012;380(9838):247257. PubMed ID: 22818937 doi:10.1016/S0140-6736(12)60646-1

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 10.

    Heath GW, Parra DC, Sarmiento OL, et al. Evidence-based intervention in physical activity: lessons from around the world. Lancet. 2012;380(9838):272281. PubMed ID: 22818939 doi:10.1016/S0140-6736(12)60816-2

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11.

    Bauman AE, Reis RS, Sallis JF, et al. Correlates of physical activity: why are some people physically active and others not? Lancet. 2012;380(9838):258271. PubMed ID: 22818938 doi:10.1016/S0140-6736(12)60735-1

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12.

    Guthold R, Stevens GA, Riley LM, Bull FC. Worldwide trends in insufficient physical activity from 2001 to 2016: a pooled analysis of 358 population-based surveys with 1·9 million participants. Lancet Glob Heal. 2018;6(10):e1077e1086. doi:10.1016/S2214-109X(18)30357-7

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13.

    Task Force on Community Preventitive Services. Recommendations to increase physical activity in communities. Am J Prev Med. 2010;22(02):6772. doi:10.1016/s0749-3797(02)00433-6

    • Search Google Scholar
    • Export Citation
  • 14.

    Haskell WL. Physical activity by self-report: a brief history and future issues. J Phys Act Health. 2012;9(suppl 1):S5S10. doi:10.1123/jpah.9.s1.s5

  • 15.

    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):181188. PubMed ID: 18091006 doi:10.1249/mss.0b013e31815a51b3

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 16.

    Powell KE, Paluch AE, Blair SN. Physical activity for health: what kind? How much? How intense? On top of what? Annu Rev Public Health. 2011;32(1):349365. PubMed ID: 21128761 doi:10.1146/annurev-publhealth-031210-101151

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17.

    Arvidsson D, Fridolfsson J, Börjesson M. Measurement of physical activity in clinical practice using accelerometers. J Intern Med. 2019;286(2):137153. PubMed ID: 30993807 doi:10.1111/joim.12908

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 18.

    Prince SA, Adamo KB, Hamel ME, Hardt J, Connor Gorber S, Tremblay M. A comparison of direct versus self-report measures for assessing physical activity in adults: a systematic review. Int J Behav Nutr Phys Act. 2008;5(1):56. PubMed ID: 18990237 doi:10.1186/1479-5868-5-56

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 19.

    Cato M, Wyka K, Ferris E, Dorn J, Thorpe L, Huang TTK. Correlates of accelerometry non-adherence in an economically disadvantaged minority urban adult population. J Sci Med Sport. 2020;23(8):746752. PubMed ID: 32085979 doi:10.1016/j.jsams.2020.01.013

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20.

    Loprinzi PD, Cardinal BJ, Crespo CJ, Brodowicz GR, Andersen RE, Smit E. Differences in demographic, behavioral, and biological variables between those with valid and invalid accelerometry data: implications for generalizability. J Phys Act Heal. 2013;10(1):7984. doi:10.1123/jpah.10.1.79

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21.

    O’Brien W, Shultz S, Firestone R, George L, Bernard B, Kruger R. Exploring the challenges in obtaining physical activity data from women using hip-worn accelerometers. Eur J Sport Med. 2017;17(7):922930. doi:10.1080/17461391.2017.1323952

    • Search Google Scholar
    • Export Citation
  • 22.

    Colley R, Gorber SC, Tremblay MS. Quality control and data reduction procedures for accelerometry-derived measures of physical activity. Heal Reports. 2010;21(1):6369.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 23.

    Krause NM, Brossard D, Scheufele DA, Xenos MA, Franke K. Americans’ trust in science and scientists. Public Opin Q. 2019;83(4):817836. doi:10.1093/poq/nfz041

    • Search Google Scholar
    • Export Citation
  • 24.

    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):197201. PubMed ID: 24297837 doi:10.1136/bjsports-2013-093154

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 25.

    Beck AM, Eyler AA, Aaron Hipp J, et al. A multilevel approach for promoting physical activity in rural communities: a cluster randomized controlled trial. BMC Public Health. 2019;19(1):110. doi:10.1186/s12889-019-6443-8

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 26.

    US Department of Agriculture. 2013 Rural-Urban Continuum Codes.

  • 27.

    Evenson KR, Sotres-Alvarez D, Deng Y, et al. Accelerometer adherence and performance in a cohort study of US Hispanic adults. Med Sci Sports Exerc. 2015;47(4):725734. PubMed ID: 25137369 doi:10.1249/MSS.0000000000000478

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 28.

    Lee PH, Macfarlane DJ, Lam TH. Factors associated with participant compliance in studies using accelerometers. Gait Posture. 2013;38(4):912917. PubMed ID: 23688408 doi:10.1016/j.gaitpost.2013.04.018

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 29.

    Roth M, Mindell J. Who provides accelerometry data? Correlates of adherence to wearing an accelerometry motion sensor: the 2008 health survey for England. J Phys Act Heal. 2012;10(1):7078. doi:10.1123/jpah.10.1.70

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30.

    Marcus BH, Forsyth LH. Motivating People to Be Physically Active. Champaign, IL: Human Kinetics; 2003:21.

  • 31.

    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(1):111. doi:10.1186/1471-2458-14-1255

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32.

    US Department of Health and Human Services. Physical Activity Guidelines for Americans. Vol. 2. Washington, DC: US Department of Health and Human Services; 2018.

    • Search Google Scholar
    • Export Citation
  • 33.

    Matthews CE, Hagströmer M, Pober DM, Bowles HR. Best practices for using physical activity monitors. Med Sci Sport Exerc. 2012;44(suppl 1):S68S76. doi:10.1249/MSS.0b013e3182399e5b

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 34.

    Evenson KR, Terry JW Jr. Assessment of differing definitions of accelerometer nonwear time. Res Q Exerc Sport. 2009;80(2):355362. PubMed ID: 19650401 doi:10.1080/02701367.2009.10599570

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 35.

    Carlson JA, Bracy NL, Sallis JF, et al. Sociodemographic moderators of relations of neighborhood safety to physical activity. Med Sci Sport Exerc. 2014;46(8):15541563. doi:10.1249/MSS.0000000000000274

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 36.

    IBM Corp. IBM SPSS Statistics for Windows. 2017.

  • 37.

    Ward DS, Evenson KR, Vaughn A, Rodgers AB, Troiano RP. Accelerometer use in physical activity: best practices and research recommendations. Med Sci Sports Exerc. 2005;37(11):S582S588. doi:10.1249/01.mss.0000185292.71933.91

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 38.

    Pew Research Center. Mobile Fact Sheet. Published 2021. https://www.pewresearch.org/internet/fact-sheet/mobile/. Accessed July 19, 2021.

  • 39.

    Hansen PG, Jespersen AM. Nudge and the manipulation of choice: a framework for the responsible use of the nudge approach to behaviour change in public policy. Eur J Risk Regul. 2013;4(1):328. doi:10.1017/S1867299X00002762

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 40.

    Bruns H, Kantorowicz-Reznichenko E, Klement K, Luistro Jonsson M, Rahali B. Can nudges be transparent and yet effective? J Econ Psychol. 2018;65(February):4159. doi:10.1016/j.joep.2018.02.002

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 41.

    Serrano N, Perez LG, Carlson J, et al. Sub-population differences in the relationship between the neighborhood environment and Latinas’ daily walking and vehicle time. J Transp Heal. 2018;8:210219. doi:10.1016/j.jth.2018.01.006

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 42.

    Perez LG, Carlson J, Slymen DJ, et al. Does the social environment moderate associations of the built environment with Latinas’ objectively-measured neighborhood outdoor physical activity? Prev Med Reports. 2016;4:551557. doi:10.1016/j.pmedr.2016.10.006

    • Crossref
    • Search Google Scholar
    • Export Citation

Gilbert, Beck, and Brownson are with the Prevention Research Center in St. Louis, Brown School at Washington University in St. Louis, St. Louis, MO, USA. Serrano is with the Institute for Health Research and Policy, University of Illinois at Chicago, Chicago, IL, USA. Brownson is also with the Division of Public Health Sciences, Department of Surgery, Washington University in St. Louis, St. Louis, MO, USA.

Gilbert (a.s.gilbert@wustl.edu) is corresponding author.
  • Collapse
  • Expand
  • 1.

    US Department of Health and Human Services. 2018 Physical Activity Guidelines Advisory Committee. Washington, DC: US Department of Health and Human Services; 2018. doi:10.1111/j.1753-4887.2008.00136.x

    • Search Google Scholar
    • Export Citation
  • 2.

    Moore SC, Lee IM, Weiderpass E, et al. Association of leisure-time physical activity with risk of 26 types of cancer in 1.44 million adults. JAMA Intern Med. 2016;176(6):816825. PubMed ID: 27183032 doi:10.1001/jamainternmed.2016.1548

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3.

    Bize R, Johnson JA, Plotnikoff RC. Physical activity level and health-related quality of life in the general adult population: a systematic review. Prev Med. 2007;45(6):401415. doi:10.1016/j.ypmed.2007.07.017

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4.

    Fan JX, Wen M, Kowaleski-Jones L. Rural–urban differences in objective and subjective measures of physical activity: findings from the national health and nutrition examination survey (NHANES) 2003–2006. Prev Chronic Dis. 2014;11:111. doi:10.5888/pcd11.140189

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5.

    Whitfield GP, Carlson SA, Ussery EN, Fulton JE, Galuska DA, Petersen R. Trends in meeting physical activity guidelines among urban and rural dwelling adults—United States, 2008–2017. MMWR Morb Mortal Wkly Rep. 2019;68(23):513518. PubMed ID: 31194722 doi:10.15585/mmwr.mm6823a1

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6.

    Cohen SA, Greaney ML, Sabik NJ. Assessment of dietary patterns, physical activity and obesity from a national survey: rural-urban health disparities in older adults. PLoS One. 2018;13(12):e0208268. doi:10.1371/journal.pone.0208268

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7.

    Singh GK, Siahpush M. Widening rural-urban disparities in all-cause mortality and mortality from major causes of death in the USA, 1969–2009. J Urban Heal. 2014;91:272292. doi:10.1007/s11524-013-9847-2

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8.

    Bolin JN, Bellamy GR, Ferdinand AO, et al. Rural healthy people 2020: new decade, same challenges. J Rural Heal. 2015;31(3):326333. doi:10.1111/jrh.12116

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9.

    Hallal PC, Andersen LB, Bull FC, et al. Global physical activity levels: surveillance progress, pitfalls, and prospects. Lancet. 2012;380(9838):247257. PubMed ID: 22818937 doi:10.1016/S0140-6736(12)60646-1

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 10.

    Heath GW, Parra DC, Sarmiento OL, et al. Evidence-based intervention in physical activity: lessons from around the world. Lancet. 2012;380(9838):272281. PubMed ID: 22818939 doi:10.1016/S0140-6736(12)60816-2

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11.

    Bauman AE, Reis RS, Sallis JF, et al. Correlates of physical activity: why are some people physically active and others not? Lancet. 2012;380(9838):258271. PubMed ID: 22818938 doi:10.1016/S0140-6736(12)60735-1

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12.

    Guthold R, Stevens GA, Riley LM, Bull FC. Worldwide trends in insufficient physical activity from 2001 to 2016: a pooled analysis of 358 population-based surveys with 1·9 million participants. Lancet Glob Heal. 2018;6(10):e1077e1086. doi:10.1016/S2214-109X(18)30357-7

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13.

    Task Force on Community Preventitive Services. Recommendations to increase physical activity in communities. Am J Prev Med. 2010;22(02):6772. doi:10.1016/s0749-3797(02)00433-6

    • Search Google Scholar
    • Export Citation
  • 14.

    Haskell WL. Physical activity by self-report: a brief history and future issues. J Phys Act Health. 2012;9(suppl 1):S5S10. doi:10.1123/jpah.9.s1.s5

  • 15.

    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):181188. PubMed ID: 18091006 doi:10.1249/mss.0b013e31815a51b3

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 16.

    Powell KE, Paluch AE, Blair SN. Physical activity for health: what kind? How much? How intense? On top of what? Annu Rev Public Health. 2011;32(1):349365. PubMed ID: 21128761 doi:10.1146/annurev-publhealth-031210-101151

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17.

    Arvidsson D, Fridolfsson J, Börjesson M. Measurement of physical activity in clinical practice using accelerometers. J Intern Med. 2019;286(2):137153. PubMed ID: 30993807 doi:10.1111/joim.12908

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 18.

    Prince SA, Adamo KB, Hamel ME, Hardt J, Connor Gorber S, Tremblay M. A comparison of direct versus self-report measures for assessing physical activity in adults: a systematic review. Int J Behav Nutr Phys Act. 2008;5(1):56. PubMed ID: 18990237 doi:10.1186/1479-5868-5-56

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 19.

    Cato M, Wyka K, Ferris E, Dorn J, Thorpe L, Huang TTK. Correlates of accelerometry non-adherence in an economically disadvantaged minority urban adult population. J Sci Med Sport. 2020;23(8):746752. PubMed ID: 32085979 doi:10.1016/j.jsams.2020.01.013

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20.

    Loprinzi PD, Cardinal BJ, Crespo CJ, Brodowicz GR, Andersen RE, Smit E. Differences in demographic, behavioral, and biological variables between those with valid and invalid accelerometry data: implications for generalizability. J Phys Act Heal. 2013;10(1):7984. doi:10.1123/jpah.10.1.79

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21.

    O’Brien W, Shultz S, Firestone R, George L, Bernard B, Kruger R. Exploring the challenges in obtaining physical activity data from women using hip-worn accelerometers. Eur J Sport Med. 2017;17(7):922930. doi:10.1080/17461391.2017.1323952

    • Search Google Scholar
    • Export Citation
  • 22.

    Colley R, Gorber SC, Tremblay MS. Quality control and data reduction procedures for accelerometry-derived measures of physical activity. Heal Reports. 2010;21(1):6369.

    • Search Google Scholar
    • Export Citation
  • 23.

    Krause NM, Brossard D, Scheufele DA, Xenos MA, Franke K. Americans’ trust in science and scientists. Public Opin Q. 2019;83(4):817836. doi:10.1093/poq/nfz041

    • Search Google Scholar
    • Export Citation
  • 24.

    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):197201. PubMed ID: 24297837 doi:10.1136/bjsports-2013-093154

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 25.

    Beck AM, Eyler AA, Aaron Hipp J, et al. A multilevel approach for promoting physical activity in rural communities: a cluster randomized controlled trial. BMC Public Health. 2019;19(1):110. doi:10.1186/s12889-019-6443-8

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 26.

    US Department of Agriculture. 2013 Rural-Urban Continuum Codes.

  • 27.

    Evenson KR, Sotres-Alvarez D, Deng Y, et al. Accelerometer adherence and performance in a cohort study of US Hispanic adults. Med Sci Sports Exerc. 2015;47(4):725734. PubMed ID: 25137369 doi:10.1249/MSS.0000000000000478

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 28.

    Lee PH, Macfarlane DJ, Lam TH. Factors associated with participant compliance in studies using accelerometers. Gait Posture. 2013;38(4):912917. PubMed ID: 23688408 doi:10.1016/j.gaitpost.2013.04.018

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 29.

    Roth M, Mindell J. Who provides accelerometry data? Correlates of adherence to wearing an accelerometry motion sensor: the 2008 health survey for England. J Phys Act Heal. 2012;10(1):7078. doi:10.1123/jpah.10.1.70

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30.

    Marcus BH, Forsyth LH. Motivating People to Be Physically Active. Champaign, IL: Human Kinetics; 2003:21.

  • 31.

    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(1):111. doi:10.1186/1471-2458-14-1255

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32.

    US Department of Health and Human Services. Physical Activity Guidelines for Americans. Vol. 2. Washington, DC: US Department of Health and Human Services; 2018.

    • Search Google Scholar
    • Export Citation
  • 33.

    Matthews CE, Hagströmer M, Pober DM, Bowles HR. Best practices for using physical activity monitors. Med Sci Sport Exerc. 2012;44(suppl 1):S68S76. doi:10.1249/MSS.0b013e3182399e5b

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 34.

    Evenson KR, Terry JW Jr. Assessment of differing definitions of accelerometer nonwear time. Res Q Exerc Sport. 2009;80(2):355362. PubMed ID: 19650401 doi:10.1080/02701367.2009.10599570

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 35.

    Carlson JA, Bracy NL, Sallis JF, et al. Sociodemographic moderators of relations of neighborhood safety to physical activity. Med Sci Sport Exerc. 2014;46(8):15541563. doi:10.1249/MSS.0000000000000274

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 36.

    IBM Corp. IBM SPSS Statistics for Windows. 2017.

  • 37.

    Ward DS, Evenson KR, Vaughn A, Rodgers AB, Troiano RP. Accelerometer use in physical activity: best practices and research recommendations. Med Sci Sports Exerc. 2005;37(11):S582S588. doi:10.1249/01.mss.0000185292.71933.91

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 38.

    Pew Research Center. Mobile Fact Sheet. Published 2021. https://www.pewresearch.org/internet/fact-sheet/mobile/. Accessed July 19, 2021.

  • 39.

    Hansen PG, Jespersen AM. Nudge and the manipulation of choice: a framework for the responsible use of the nudge approach to behaviour change in public policy. Eur J Risk Regul. 2013;4(1):328. doi:10.1017/S1867299X00002762

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 40.

    Bruns H, Kantorowicz-Reznichenko E, Klement K, Luistro Jonsson M, Rahali B. Can nudges be transparent and yet effective? J Econ Psychol. 2018;65(February):4159. doi:10.1016/j.joep.2018.02.002

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 41.

    Serrano N, Perez LG, Carlson J, et al. Sub-population differences in the relationship between the neighborhood environment and Latinas’ daily walking and vehicle time. J Transp Heal. 2018;8:210219. doi:10.1016/j.jth.2018.01.006

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 42.

    Perez LG, Carlson J, Slymen DJ, et al. Does the social environment moderate associations of the built environment with Latinas’ objectively-measured neighborhood outdoor physical activity? Prev Med Reports. 2016;4:551557. doi:10.1016/j.pmedr.2016.10.006

    • Crossref
    • Search Google Scholar
    • Export Citation
All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 3823 407 125
PDF Downloads 745 164 39