Exploring the challenges in obtaining physical activity data from women using hip-worn accelerometers

Association of sociodemographic, anthropometric, and sleep quality factors with accelerometer-measured sitting and physical activity times among Emirati working women during the COVID-19 pandemic: A cross-sectional study

BACKGROUND

Although a significant lack of physical activity has been linked to an increase in obesity among Emirati women, the factors associated with accelerometer-measured sitting and physical activity times in Emirati women remain unclear.

OBJECTIVES

To explore the association of accelerometer-measured sitting and physical activity times with sociodemographic, anthropometric, and sleep quality factors among Emirati working women.

DESIGN

A cross-sectional study.

METHODS

A convenience sample of 163 healthy working Emirati women aged 18-45 years was used. Sedentary and physical activity times were measured using the Fibion accelerometers worn on the thighs for 4-7 days. General demographic information, anthropometric measurements, and self-reported sleep quality (Pittsburgh sleep quality index score) were collected. Only participants who had valid data (i.e. wear time of ⩾600 min (10 h) per day for a minimum of 4-7 days) were evaluated. All values were normalized to a 16-h day to mitigate differences in wear time among the participants.

RESULTS

Overall, 110 Emirati women were included. The mean sitting time per 16-h day was 11.6 ± 1.1 h; mean moderate activity time per day, 40.88 ± 17.99 min; and mean vigorous activity time per day, 2.41 ± 1.21 min. Longer sitting time was associated with high body fat, secondary education, and divorce. Sitting time was reduced in those with good sleep quality. Moderate-to-vigorous physical activity time was increased in women with postgraduate education and was decreased in women with a longer work experience and with comorbidities. The total activity time increased with increasing age and good sleep quality, whereas it decreased with increasing body fat, presence of at least one comorbidity, secondary education, and divorce.

CONCLUSION

Certain demographic, anthropometric, and sleep quality factors were associated with accelerometer-measured sitting and physical activity times among Emirati working women. Future longitudinal studies should consider these factors when investigating predictors of physical activity levels in this population.

Monitoring physical activity using wearable technology in people with Achilles tendinopathy undergoing physiotherapy treatment: A feasibility prospective cohort study

OBJECTIVES

Physical activity modification is an important part of the management of Achilles tendinopathy. However, to our knowledge, there is a lack of evidence on objective physical activity assessment in Achilles tendinopathy. The purpose of this study is to (1) assess feasibility of using an inertial measurement unit (IMU) to monitor physical activity and IMU-derived biomechanical measures over 12-week treatment course by a physiotherapist; (2) conduct a preliminary analysis of changes in physical activity over 12-weeks.

DESIGN

A feasibility prospective cohort study SETTING: A community setting.

PARTICIPANTS

People with Achilles tendinopathy who had recently commenced (≤2 sessions), or were about to commence, treatment with a physiotherapist MAIN OUTCOME MEASURES: Participants wore a shank-mounted IMU on the affected side for one week at baseline, 6-, and 12-week follow-ups. The outcomes were pain/symptom severity, IMU-derived physical activity and biomechanical measures (stride rate, peak shank angular velocity, and peak shank acceleration).

RESULTS

Thirty participants were recruited. There was a high retention rate (97%), response rate (97%), and IMU wear compliance at each timepoint (>93%). For pain/symptom severity, a significant time effect was observed between baseline and 12-week follow-up. Physical activity and IMU-derived biomechanical measures did not change over 12 weeks. Physical activity decreased at the 6-week follow-up but only returned to the baseline level at 12-week follow-up.

CONCLUSIONS

A larger-scale cohort study assessing clinical outcomes and physical activity appears feasible. Preliminary data indicate that physical activity may not change significantly over 12-weeks in people undergoing physiotherapy management for Achilles tendinopathy. CONTRIBUTION OF THE PAPER.

The Challenge of Incomplete Data in Accelerometer Studies: Characteristics of Nonparticipation and Noncompliance in a Nationwide Sample of Adolescents and Young Adults in Germany

BACKGROUND

Incomplete data due to nonparticipation and noncompliance with the study protocol can bias the results of studies. We investigated how a nationwide accelerometer sample of adolescents and young adults is affected by such incomplete data.

METHODS

We analyzed cross-sectional data from 6465 participants (11-31 y old) who participated in a national health survey in Germany (KiGGS Wave 2; 2014-2017). The data included information about the participation in the measurement of physical activity using accelerometers, compliance with the wear-time protocol, and sociodemographic and health-related variables. Multivariable regression analyses were conducted to detect factors associated with incomplete data.

RESULTS

Of the total sample, 78.0% participated in the accelerometer part of the study, and 83.5% of the participants with data available complied with the wear-time protocol. In 11- to 17-year-olds, the likelihood of having incomplete accelerometer data was higher in boys, older adolescents, adolescents with a lower sociodemographic status, adolescents with overweight, adolescents not participating in organized sport, adolescents not speaking only German at home, current smokers, and adolescents having a higher soft drink consumption. In 18- to 31-year-olds, the likelihood of having incomplete accelerometer data was higher in men, adults with a lower educational level, adults not speaking only German at home, and adults who smoke.

CONCLUSIONS

Our results suggest that accelerometer samples are biased such that participants with more beneficial health behaviors provide complete accelerometer data more often. This knowledge should be used to design effective recruitment strategies and should be considered when interpreting results of accelerometer studies.

How many days of indoor positioning system data are required to characterise typical movement behaviours of office workers?

Indoor Positioning Systems (IPS) appear to offer great potential to study the movement and interaction of people and their working environment, including office workplaces. But little is known about appropriate durations for data collection. In this study, location observations collected from 24 office workers on a 1220 m² office floor over a 3-month period, were analysed to determine how many days are required to estimate their typical movement and spatial behaviours. The analysis showed that up to 8 days of data was sufficient to characterise participants' typical daily movement behaviours and 10 days were required to estimate their typical spatial mobility. However, the results also indicate that 5 weeks of data collection are required to gather the necessary 10 days of data from each participant. These findings will help researchers and workplace professionals to understand the capabilities and requirements of IPS when considering their use in indoor work environments.

Physical activity and pregnancy outcomes: an expert review

This expert review aimed to assess current literature on the effect and tracking of physical activity during pregnancy and associated outcomes. Self-reported physical activity may be inaccurate given the subjective nature of the questionnaires. The accelerometer ActiGraph is considered the "gold standard" to objectively measure physical activity. However, other more user-friendly wearable devices are now widely available and may accurately track physical activity. Conclusive data from both validated activity questionnaires and accelerometers indicate that physical activity is safe during pregnancy. In addition, studies of physical activity during pregnancy that evaluate pregnancy outcomes have found reduced risks of preterm birth, preeclampsia, and gestational diabetes mellitus and improved mental health among individuals who regularly engage in physical activity. In the United States, approximately 48% of pregnant individuals gain more than the recommended amount of weight during pregnancy; excessive gestational weight gain is associated with an increased risk of maternal and fetal complications, including preterm birth, preeclampsia, and gestational diabetes mellitus, and corresponding higher adverse short- and long-term maternal and offspring health outcomes. Although physical activity is safe during pregnancy and may reduce excessive gestational weight gain and resultant pregnancy complications, further research is needed to determine the frequency and duration of specific types of physical activity during pregnancy. Providers should encourage physical activity before and during pregnancy and educate patients regarding the benefits and safety of physical activity.

Replacing Sedentary Time with Physically Active Behaviour Predicts Improved Body Composition and Metabolic Health Outcomes

Background: Discretionary leisure time for health-promoting physical activity (PA) is limited. This study aimed to predict body composition and metabolic health marker changes from PA reallocation using isotemporal substitution analysis. Methods: Healthy New Zealand women (n = 175; 16–45 y) with high BMI (≥25 kg/m2) and high body fat percentage (≥30%) were divided into three groups by ethnicity (Māori n = 37, Pacific n = 54, and New Zealand European n = 84). PA, fat mass, lean mass, and metabolic health were assessed. Isotemporal substitution paradigms reallocated 30 min/day of sedentary behaviour to varying PA intensities. Results: Reallocating sedentary behaviour with moderate intensity, PA predicted Māori women would have improved body fat% (14.83%), android fat% (10.74%), and insulin levels (55.27%) while the model predicted Pacific women would have improved waist-to-hip (6.40%) and android-to-gynoid (19.48%) ratios. Replacing sedentary time with moderate-vigorous PA predicted Māori women to have improved BMI (15.33%), waist circumference (9.98%), body fat% (16.16%), android fat% (12.54%), gynoid fat% (10.04%), insulin (55.58%), and leptin (43.86%) levels; for Pacific women, improvement of waist-to-hip-ratio (5.30%) was predicted. Conclusions: Sedentary behaviour must be substituted with PA of at least moderate intensity to reap benefits. Māori women received the greatest benefits when reallocating PA. PA recommendations to improve health should reflect the needs and current activity levels of specific populations.

Characterisation of Temporal Patterns in Step Count Behaviour from Smartphone App Data: An Unsupervised Machine Learning Approach

The increasing ubiquity of smartphone data, with greater spatial and temporal coverage than achieved by traditional study designs, have the potential to provide insight into habitual physical activity patterns. This study implements and evaluates the utility of both K-means clustering and agglomerative hierarchical clustering methods in identifying weekly and yearlong physical activity behaviour trends. Characterising the demographics and choice of activity type within the identified clusters of behaviour. Across all seven clusters of seasonal activity behaviour identified, daylight saving was shown to play a key role in influencing behaviour, with increased activity in summer months. Investigation into weekly behaviours identified six clusters with varied roles, of weekday versus weekend, on the likelihood of meeting physical activity guidelines. Preferred type of physical activity likewise varied between clusters, with gender and age strongly associated with cluster membership. Key relationships are identified between weekly clusters and seasonal activity behaviour clusters, demonstrating how short-term behaviours contribute to longer-term activity patterns. Utilising unsupervised machine learning, this study demonstrates how the volume and richness of secondary app data can allow us to move away from aggregate measures of physical activity to better understand temporal variations in habitual physical activity behaviour.

The validity of the 'General Practice Physical Activity Questionnaire' against accelerometery in patients with chronic kidney disease

Background: The majority of patients with chronic kidney disease (CKD) are physically inactive. Simple yet accurate assessment of physical activity is important in identifying those in need of intervention. The 'General Practice Physical Activity Questionnaire' (GPPAQ) is a well-used clinical and research tool, but has not been validated.Methods: Forty individuals with CKD (age 62.5 (SD: 11.1) years, estimated glomerular filtration rate (eGFR) 33.2 (SD:19.1) ml/min/1.73 m²) completed the GPPAQ and objective physical activity was measured using a GENEActiv accelerometer for 7 days. Physical activity status was grouped as 'Active' (i.e. meeting current physical activity UK guidelines) or 'Inactive.' Sensitivity and specificity were calculated. Accuracy was defined as the probability the GPPAQ could correctly classify a patient as either 'Active' or 'Inactive' (based on accelerometery).Results: Using accelerometery, 18% of participants met the current UK guidelines, whereas 27% were classed as 'Active' according to GPPAQ. Sensitivity of the GPPAQ was 54.6% and specificity was 96.6%. The 'accuracy' of the GPPAQ was 85.0%. The accuracy of the GPPAQ was greater in males and those not in employment/retired, although these differences were not statistically significant.Conclusions: The GPPAQ may be a useful tool to identify CKD patients who would benefit most from a physical activity intervention. In particular, the GPPAQ can accurately identify those not sufficiently active.

Process evaluation of the 'Singapore Physical Activity and Nutrition Study'

INTRODUCTION

The Singapore Physical Activity and Nutrition Study (SPANS) aimed to improve the physical activity (PA) and nutrition behaviours of Singaporean women aged 50 years and over. The SPANS program consisted of PA classes, nutrition workshops, telephone dietary counselling, health booklets, a health calendar and program ambassadors. This study aimed to assess and understand the implementation of the program strategies and gain insight into process evaluation components to inform future programs.

METHODS

The evaluation was guided by a process evaluation framework and collected data via questionnaires (n = 209), program ambassador documentation and exit interviews with program completers (n = 13) and non-completers (n = 12).

RESULTS

In total, 295 participants completed the program (response rate = 84 %). Participants reported high levels of satisfaction with the overall program (99.5 %) and program activities (96.7 %), and also rated program ambassadors highly. Participation rates were highest for telephone dietary counselling sessions. The main reason for not attending program activities was having a 'busy schedule' (n = 158). Participants cited a need for improved recreational centre facilities and increased flexibility around program delivery.

CONCLUSIONS

The process evaluation showed that the program strategies were implemented as planned and were deemed suitable for supporting behaviour change among Singaporean women aged 50 years and over. The program reached and involved the majority of participants throughout the six months. The combination of practical educational resources and supportive program ambassadors were key strategies that facilitated positive PA and dietary behaviours. However, there needs to be some flexibility in the delivery of programs. The findings of this research may inform other programs in the region.

Demographic-specific Validity of the Cancer Prevention Study-3 Sedentary Time Survey

PURPOSE

This study examined the 1-yr test-retest reliability and criterion validity of sedentary time survey items in a subset of participants from a large, nationwide prospective cohort.

METHODS

Participants included 423 women and 290 men age 31 to 72 yr in the Cancer Prevention Study-3. Reliability was assessed by computing Spearman correlation coefficients between responses from prestudy and poststudy surveys. Validity was assessed by comparing survey-estimated sedentary time with a latent variable representing true sedentary time estimated from the 7-d diaries, accelerometry, and surveys through the method of triads. Sensitivity analyses were restricted to 566 participants with an average of 14+ h of diary and accelerometer data per day for 7 d per quarter.

RESULTS

Reliability estimates for total sitting time were moderate or strong across all demographic strata (Spearman ρ ≥ 0.6), with significant differences by race (P = 0.01). Reliability estimates were strongest for the TV-related sedentary time item (Spearman ρ, 0.74; 95% confidence interval, 0.70-0.77). The overall validity coefficient (VC) for survey-assessed total sedentary time was 0.62 (95% confidence interval, 0.55-0.69), although VC varied by age group and activity level (P < 0.05). However, VC were similar across groups (P < 0.05) when restricting to highly compliant participants in a sensitivity analysis.

CONCLUSIONS

The Cancer Prevention Study-3 sedentary behavior questionnaire has acceptable reliability and validity for ranking or categorizing participants according to sedentary time. Acceptable reliability and validity estimates persist across various demographic subgroups.

Promoting physical activity among community groups of older women in socio-economically disadvantaged areas: randomised feasibility study

BACKGROUND

Insufficient physical activity (PA) is a major public health issue. Whilst PA is an important contributor to disease prevention, engagement in PA decreases with age, particularly among women in socio-economically disadvantaged areas. Research using existing support networks to engage 'hard to reach' populations in PA interventions is sparse. We developed and tested the feasibility of a PA-promoting intervention for older women within existing community groups in socio-economically disadvantaged areas.

METHODS

The Medical Research Council guidelines for complex interventions were used to guide the intervention's development. We recruited participants (n = 40) from older (aged ≥50 years) women's groups from four different community centres. A 12-week programme was delivered during existing sessions, informed by Social Practice Theory. The sessions provided education about PA, social support in the form of a PA 'buddy', group discussion and follow-up telephone calls, as well as printed information about local opportunities to participate in PA. The main uncertainties tested were rates of participant recruitment, retention, and completion of assessments of PA by accelerometry and of mental health using the Hospital Anxiety and Depression Scale (HADS). Intervention acceptability was assessed by questionnaire, and focus group interviews elicited participants' views about the intervention. Qualitative data were subjected to framework analysis.

RESULTS

The recruitment rate was high; 87% (n = 40/46) of women consented to participate, and 78% (n = 31) attended all education sessions. Uptake of follow-up telephone calls and PA 'buddies' was low. Few participants provided valid accelerometer data, but 63% (n=25) completed the HADS questionnaire at all time points. The printed materials and education sessions were viewed positively; telephone calls and 'buddy' support were not valued. Participants believed that organised group activities would lead to increased PA engagement, and whilst participants disliked wearing a waist accelerometer, they thought that regular PA feedback would facilitate necessary goal-setting.

CONCLUSIONS

High recruitment and retention rates suggest that use of existing social support groups is an acceptable and attractive method of delivering a PA intervention to this population. A randomised controlled trial of the intervention appears feasible, but its design requires refinement of the social support component, facilitation of goal-setting and reconsideration of the assessment of PA.

TRIAL REGISTRATION

ClinicalTrials.gov, NCT02880449 . Registered on 26 August 2016.

Device-measured physical activity data for classification of patients with ventricular arrhythmia events: A pilot investigation

Low levels of physical activity are associated with increased mortality risk, especially in cardiac patients, but most studies are based on self-report. Cardiac implantable electronic devices (CIEDs) offer an opportunity to collect data for longer periods of time. However, there is limited agreement on the best approaches for quantification of activity measures due to the time series nature of the data. We examined physical activity time series data from 235 subjects with CIEDs and at least 365 days of uninterrupted measures. Summary statistics for raw daily physical activity (minutes/day), including statistical moments (e.g., mean, standard deviation, skewness, kurtosis), time series regression coefficients, frequency domain components, and forecasted predicted values, were calculated for each individual, and used to predict occurrence of ventricular tachycardia (VT) events as recorded by the device. In unsupervised analyses using principal component analysis, we found that while certain features tended to cluster near each other, most provided a reasonable spread across activity space without a large degree of redundancy. In supervised analyses, we found several features that were associated with the outcome (P < 0.05) in univariable and multivariable approaches, but few were consistent across models. Using a machine-learning approach in which the data was split into training and testing sets, and models ranging in complexity from simple univariable logistic regression to ensemble decision trees were fit, there was no improvement in classification of risk over naïve methods for any approach. Although standard approaches identified summary features of physical activity data that were correlated with risk of VT, machine-learning approaches found that none of these features provided an improvement in classification. Future studies are needed to explore and validate methods for feature extraction and machine learning in classification of VT risk based on device-measured activity.