Accelerometer-derived physical activity in those with cardio-metabolic disease compared to healthy adults: a UK Biobank study of 52,556 participants

Accelerometer-derived physical activity in those with cardio-metabolic disease compared to... Aim Cardio-metabolic disease and physical activity are closely related but large-scale objective studies which measure physical activity are lacking. Using the largest accelerometer cohort to date, we aimed to investigate whether there is an association between disease status and accelerometer variables after a 5-year follow-up. Methods 106,053 UK Biobank participants wore a wrist-worn GENEactiv monitor. Those with acceptable wear time (> 3 days) were split into 4 cardio-metabolic disease groups based on self-report disease status which was collected 5 ± 1 years prior. Multiple linear regression models were used to investigate associations, controlling for confounders and stratified for gender. Results Average daily acceleration was lower in men (‘healthy’-42 ± 15 mg v ‘Type 2 diabetes + cardiovascular disease (CVD)’-31 ± 12 mg) and women (‘healthy’-44 ± 13 mg v ‘Type 2 diabetes + CVD’-31 ± 11 mg) with cardio-metabolic dis- ease and this was consistent across both week and weekend days. Men and women with the worst cardio-metabolic disease perform around half of moderate to vigorous physical activity on a daily basis compared to healthy individuals, and spend almost 7 h per day in 30 min inactivity bouts. Significant associations were seen between cardio-metabolic disease and accelerometer variables 5 years on when controlling for confounders. Conclusion In the largest accelerometer cohort to date, there are significant associations between cardio-metabolic disease and physical activity variables after 5 years of follow-up. Triaxial accelerometers provide enhanced measurement opportuni- ties for measuring lifestyle behaviours in chronic disease. Keywords Accelerometer · Physical activity · Cardiovascular disease · Type 2 diabetes Abbreviations CVD Cardiovascular disease METs Metabolic equivalents Managed by Antonio Secchi. mg Milligravity Harley Fuller: joint first author. MVPA Moderate to vigorous physical activity * Sophie Cassidy sophie.cassidy@newcastle.ac.uk Introduction Clinical Exercise Research Group, Institute of Cellular Medicine, Faculty of Medical Sciences, Newcastle Physical activity is closely associated with cardio-metabolic University, 4th Floor William Leech Building, health [1], therefore accurate assessment is vital. Objective Newcastle upon Tyne NE2 4HH, UK measures of physical activity have become widespread and Prevention Research Collaboration, Sydney School of Public offer numerous advantages over self-report methods. Over Health, Charles Perkins Centre D17, Level 6 The Hub, the past decade, there has been a shift in using hip-worn uni- University of Sydney, Sydney, NSW 2006, Australia 3 axial accelerometers, to wrist-worn triaxial accelerometers Institute of Neuroscience, Faculty of Medical Sciences, which continuously sample and store raw acceleration. These Newcastle University, Newcastle upon Tyne NE2 4HH, UK 4 offer improved precision and enhanced measurement oppor - NIHR Innovation Observatory, Newcastle University, tunities, but large cohorts with triaxial accelerometer are Newcastle Upon Tyne, UK Vol.:(0123456789) 1 3 Acta Diabetologica lacking. Prior to the UK Biobank, the largest cohort of indi- performed using SAS OnDemand for Academics (SAS Insti- viduals with these devices included around 20,000 individu- tute, Care, North Carolina, USA) software. als [2]. This highlights the scale of the UK Biobank Study, whereby data was collected and analysed in > 100,000 par- ticipants—making it the world’s largest objective study of its Results kind. We used the UK Biobank, to investigate whether there is an association between cardio-metabolic disease status and 103,578 datasets were received, but only 52,556 fit into accelerometer variables after a 5-year follow-up. the four disease groups and were analysed. Those excluded reported a wide range of other diseases including respira- tory, gastrointestinal, renal, neurology, musculoskeletal, Materials and methods haematology, gynaecology, immunological and infectious. Those with cardio-metabolic disease were less active dur- UK Biobank baseline assessments occurred between 2007 ing waking hours, demonstrated by a reduction in daytime and 2010, when the following covariate data was collected; acceleration (mg) in both men and women, which was con- age, BMI, Townsend Deprivation Index, ethnicity, smok- sistent in both week and weekend days (Table 1). Even in ing and alcohol status, fruit and vegetable consumption, the most active 5 h of the day, those with cardio-metabolic self-report weekly moderate to vigorous physical activity disease performed a lower intensity of activity compared (MVPA). Self-report disease status was used to define four to those with no disease. During waking hours, total time disease groups spanning cardio-metabolic health, which spent in each of the activity thresholds (light time, moderate included ‘Healthy’ (individuals who reported no disease), time and vigorous time) declined across cardio-metabolic ‘Cardiovascular disease (CVD)’, Type 2 diabetes minus disease groups. Both men and women with ‘Type 2 diabe- CVD, and ‘Type 2 diabetes + CVD’. tes + CVD’ performed half the level of MVPA when con- Between February 2013 and December 2015, a subset of sidered as MVPA1min or MVPA10min bouts. Inactive time individuals was invited to wear an Axivity AX3 wrist-worn was higher across all cardio-metabolic groups and a similar triaxial accelerometer for 7 consecutive days. Raw accelerom- pattern was observed for Inactivity30min, whereby those eter data were processed using GGIR V1.5-9 package (R Core with ‘Type 2 diabetes + CVD’ spent almost 7 h of the day in Team, Vienna, Austria) (https://cr an.rroject.or g/web/packag es/ 30 min inactivity bouts (Inactivity30min). GGIR/inde x .html) [3]. We defined MVPA using a 100 mil- Figure 1 shows the prospective associations of baseline ligravity (mg) cut-off, based on laboratory findings [4 ]. Simi- disease status with objective physical activity variables after larly, 6 metabolic equivalents (METs) is classified as vigorous 5 years of follow-up, when adjusting for confounders such as activity and was equivalent to an acceleration around 400 mg BMI. There were significant inverse associations of disease [4]. ‘Light’ (< 3 METs) was defined as anything between 40 and status with average acceleration across week and weekend 100 mg, and ‘Inactivity’ as anything below 40 mg [2]. Within days for both men and women. Similar patterns were seen each threshold, total activity time within waking hours was with bouted and unbouted MVPA, whereas disease status calculated (Light time, Moderate time, Vigorous time, Inactiv- demonstrated a slightly weaker positive association between ity time). Additionally, time spent in 1–5 min (MVPA1min) bouted and unbouted inactivity. Sensitivity analysis demon- and 10 min (MVPA10min) bouts of MVPA and time spent in strated similar associations. 30 min of inactivity (Inactivity30mins) was calculated. Discussion Statistical analysis In the largest objective cohort to date, there is a decrease in Due to the large sample size, any small difference in accel- objectively measured physical activity from healthy to CVD eration mean was significant, therefore these results are to diabetes patients which was consistent whether week or not reported. Multiple linear regression models were used weekend acceleration was measured, or bouted or unbouted to investigate the association between cardio-metabolic activity was used. These results demonstrate the usefulness disease and objective physical activity, after adjusting for of accelerometers in exploring the relationship between baseline covariates. Physical activity variables did not meet physical activity and chronic disease. The results also pro- assumptions of normality, therefore were transformed. All vide a platform for future novel explorations including the analyses were stratified by gender due to significant interac- temporal distribution and patterns of physical activity, sleep tion between gender and outcome variables. To determine and sedentary behaviour. the robustness of the results, sensitivity analysis was per- These devices make it possible to measure bouted and formed with a more stringent cut-point of 120 and 50 mg unbouted activity. We chose to focus on MVPA bouts in to define MVPA and inactivity, respectively. Analyses were 1 3 Acta Diabetologica Table 1 Physical activity acceleration values from Axivity in all participants (n = 52,424) according to disease status and stratified for gender Male (n = 24,880) Female (n = 27,544) Healthy CVD Type 2 Type 2 dia- Healthy CVD Type 2 Type 2 dia- (n = 11,232) (n = 11,996) diabetes betes + CVD (n = 14,960) (n = 11,746) diabetes betes + CVD minus CVD (n = 1218) minus CVD (n = 561) (n = 434) (n = 277) Age, years (SD) 54.3 (8.0) 59.6 (6.8) 59.1 (6.9) 61.0 (5.9) 53.6 (7.6) 58.5 (7.0) 58.6 (6.2) 59.9 (6.5) BMI, kg/m (SD) 26.3 (3.5) 28.4 (4.2) 29.5 (4.4) 31.4 (5.3) 25.2 (4.1) 28.0 (5.4) 31.4 (6.5) 33.2 (6.4) Physical activity  Average accel- eration values, mg (SD)   Daytime 42 (15) 36 (12) 34 (11) 31 (12) 44 (13) 38 (12) 35 (12) 31 (11) acceleration   Acceleration 0.63 (1.04) 0.69 (1.10) 0.69 (0.74) 0.79 (0.93) 0.54 (0.94) 0.57 (0.73) 0.70 (1.36) 0.72 (0.82) for least active 5 h   Acceleration 67 (28) 56 (22) 52 (18) 47 (25) 66 (23) 57 (18) 51 (18) 46 (17) for most active 5 h   Weekday 30 (6) 26 (8) 24 (7) 22 (11) 30 (8) 27 (8) 24 (7) 22 (7) acceleration across night and day   Weekend 30 (12) 25 (10) 23 (8) 21 (7) 30 (10) 26 (8) 23 (8) 21 (7) acceleration across night and day  Total time spent across differ - ent thresholds during waking time (min/ day)   Inactivity 588 (75) 604 (3) 615 (78) 624 (74) 568 (77) 583 (75) 599 (73) 617 (80) time   Light time 162 (47) 156 (47) 153 (50) 146 (48) 182 (46) 178 (48) 167 (55) 156 (53)   Moderate 96 (45) 79 (40) 72 (39) 61 (38) 104 (44) 87 (43) 75 (43) 62 (40) time   Vigorous time 6.12 (7.6) 3.58 (5.04) 2.64 (3.01) 1.87 (2.50) 4.8 (6.3) 2.8 (3.8) 2.1 (3.3) 1.5 (2.5)  Bouts of activ- ity during waking time (min/day)   MVPA10min 22 (28) 14 (20) 13 (19) 8 (19) 20 (25) 12 (19) 9 (17) 5 (12)   MVPA1min 23 (14) 18 (12) 16 (12) 13 (11) 25 (14) 20 (13) 16 (13) 13 (12)   Inactivity- 357 (124) 394 (128) 412 (134) 432 (133) 318 (115) 353 (122) 380 (139) 419 (141) 30min regression models, as it has been previously shown that bouts this reason, identifying associations with > 1 min bouts of of MVPA have stronger associations with metabolic health, MVPA is the most informative. Inactivity levels were high compared to unbouted activity [5]. Time spent in unbouted regardless of disease status, however in this study, associa- activity seems higher than what would be expected, with an tions between inactivity and cardio-metabolic were not as average of around 60 min/day for those with ‘Type 2 diabetes strong as other accelerometer variables. This most likely and CVD’, but this could capture sporadic arm movements reflects the methodological limitations in defining ‘inactiv - which cannot be separated from true physical activity. For ity’ using accelerometers. Due to the postural component, 1 3 Acta Diabetologica Fig. 1 Associations of disease group at baseline with objective were adjusted for age, BMI, Townsend Deprivation Index, ethnic- physical activity after an average 5 ± 1  years of follow-up. Weekday ity, smoking, fruit and vegetable intake, alcohol, self-report weekly acceleration and Inactivity30min were log transformed, and week- MVPA, follow-up time (Women = solid line + triangle, Men = dotted end acceleration and MVPA1min were log+1 transformed. Models line + square). T2D type 2 diabetes it is currently difficult to distinguish between light physical Compliance with ethical standards activity, e.g., standing, and sedentary behaviour (i.e., reclin- Conflict of interest MIT is a founder of Changing Health Ltd, a digital ing/sitting), but efforts are underway to validate and define education company. SC, HF, JC, MC, AB declare that they have no con- this behaviour using accelerometers, as well as sleep. An flict of interest. important limitation with this study is that the direction of causality cannot be distinguished. Human and animal rights All procedures followed were in accordance with the ethical standards of the responsible committee on human experi- Overall, strong and consistent relationships between mentation (institutional and national) and with the Helsinki Declaration cardio-metabolic disease and triaxial accelerometry, dem- of 1975, as revised in 2008 (5). onstrate enhanced measurement opportunities and greater insights going forward. Informed consent Informed consent was obtained from all patients for being included in the study. Acknowledgements This research was conducted using the UK Biobank resource. The authors would like to thank the UK Biobank Open Access This article is distributed under the terms of the Crea- participants and investigators for making this study possible. We would tive Commons Attribution 4.0 International License (http://creat iveco also like to acknowledge support from the Newcastle University Centre mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- for Ageing and Vitality sponsored by the BBSRC, EPSRC, ESRC and tion, and reproduction in any medium, provided you give appropriate MRC for providing support. credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Author contribution statement MIT and SC designed the study analy- sis. SC, HF, JYC, MC, AB, and MIT assisted in data analysis and References interpretation. SC and HF wrote the manuscript. All authors critically reviewed the manuscript and approved the final version for publication. ‘MIT is the guarantor of this work’. 1. Healy GN et al (2008) Objectively measured sedentary time, physi- cal activity, and metabolic risk: the Australian Diabetes, Obesity and Lifestyle Study (AusDiab). Diabetes Care 31(2):369–371 Funding MIT was supported by a Senior Research Fellowship from 2. Rowlands AV et al (2017) Accelerometer-assessed physical activity the National Institute for Health Research. in epidemiology. Med Sci Sport Exerc 4:1 1 3 Acta Diabetologica 3. Doherty A et al (2017) Large scale population assessment of physi- 5. Strath SJ, Holleman RG, Ronis DL, Swartz AM, Richardson CR cal activity using wrist worn accelerometers: the UK Biobank (2008) Objective physical activity accumulation in bouts and study. PLoS One 12(2):1–14. https://doi.org/10.1371/journal. nonbouts and relation to markers of obesity in US adults. Prev pone.0169649 Chronic Dis 5(4):A131 4. Hildebrand M, Van Hees VT, Hansen BH, Ekelund U (2014) Age group comparability of raw accelerometer output from wrist- and hip-worn monitors. Med Sci Sport Exerc 46(9):1816–1824 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Acta Diabetologica Springer Journals

Accelerometer-derived physical activity in those with cardio-metabolic disease compared to healthy adults: a UK Biobank study of 52,556 participants

Free
5 pages
Loading next page...
 
/lp/springer_journal/accelerometer-derived-physical-activity-in-those-with-cardio-metabolic-muPF4dl0q1
Publisher
Springer Milan
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Internal Medicine; Diabetes; Metabolic Diseases
ISSN
0940-5429
eISSN
1432-5233
D.O.I.
10.1007/s00592-018-1161-8
Publisher site
See Article on Publisher Site

Abstract

Aim Cardio-metabolic disease and physical activity are closely related but large-scale objective studies which measure physical activity are lacking. Using the largest accelerometer cohort to date, we aimed to investigate whether there is an association between disease status and accelerometer variables after a 5-year follow-up. Methods 106,053 UK Biobank participants wore a wrist-worn GENEactiv monitor. Those with acceptable wear time (> 3 days) were split into 4 cardio-metabolic disease groups based on self-report disease status which was collected 5 ± 1 years prior. Multiple linear regression models were used to investigate associations, controlling for confounders and stratified for gender. Results Average daily acceleration was lower in men (‘healthy’-42 ± 15 mg v ‘Type 2 diabetes + cardiovascular disease (CVD)’-31 ± 12 mg) and women (‘healthy’-44 ± 13 mg v ‘Type 2 diabetes + CVD’-31 ± 11 mg) with cardio-metabolic dis- ease and this was consistent across both week and weekend days. Men and women with the worst cardio-metabolic disease perform around half of moderate to vigorous physical activity on a daily basis compared to healthy individuals, and spend almost 7 h per day in 30 min inactivity bouts. Significant associations were seen between cardio-metabolic disease and accelerometer variables 5 years on when controlling for confounders. Conclusion In the largest accelerometer cohort to date, there are significant associations between cardio-metabolic disease and physical activity variables after 5 years of follow-up. Triaxial accelerometers provide enhanced measurement opportuni- ties for measuring lifestyle behaviours in chronic disease. Keywords Accelerometer · Physical activity · Cardiovascular disease · Type 2 diabetes Abbreviations CVD Cardiovascular disease METs Metabolic equivalents Managed by Antonio Secchi. mg Milligravity Harley Fuller: joint first author. MVPA Moderate to vigorous physical activity * Sophie Cassidy sophie.cassidy@newcastle.ac.uk Introduction Clinical Exercise Research Group, Institute of Cellular Medicine, Faculty of Medical Sciences, Newcastle Physical activity is closely associated with cardio-metabolic University, 4th Floor William Leech Building, health [1], therefore accurate assessment is vital. Objective Newcastle upon Tyne NE2 4HH, UK measures of physical activity have become widespread and Prevention Research Collaboration, Sydney School of Public offer numerous advantages over self-report methods. Over Health, Charles Perkins Centre D17, Level 6 The Hub, the past decade, there has been a shift in using hip-worn uni- University of Sydney, Sydney, NSW 2006, Australia 3 axial accelerometers, to wrist-worn triaxial accelerometers Institute of Neuroscience, Faculty of Medical Sciences, which continuously sample and store raw acceleration. These Newcastle University, Newcastle upon Tyne NE2 4HH, UK 4 offer improved precision and enhanced measurement oppor - NIHR Innovation Observatory, Newcastle University, tunities, but large cohorts with triaxial accelerometer are Newcastle Upon Tyne, UK Vol.:(0123456789) 1 3 Acta Diabetologica lacking. Prior to the UK Biobank, the largest cohort of indi- performed using SAS OnDemand for Academics (SAS Insti- viduals with these devices included around 20,000 individu- tute, Care, North Carolina, USA) software. als [2]. This highlights the scale of the UK Biobank Study, whereby data was collected and analysed in > 100,000 par- ticipants—making it the world’s largest objective study of its Results kind. We used the UK Biobank, to investigate whether there is an association between cardio-metabolic disease status and 103,578 datasets were received, but only 52,556 fit into accelerometer variables after a 5-year follow-up. the four disease groups and were analysed. Those excluded reported a wide range of other diseases including respira- tory, gastrointestinal, renal, neurology, musculoskeletal, Materials and methods haematology, gynaecology, immunological and infectious. Those with cardio-metabolic disease were less active dur- UK Biobank baseline assessments occurred between 2007 ing waking hours, demonstrated by a reduction in daytime and 2010, when the following covariate data was collected; acceleration (mg) in both men and women, which was con- age, BMI, Townsend Deprivation Index, ethnicity, smok- sistent in both week and weekend days (Table 1). Even in ing and alcohol status, fruit and vegetable consumption, the most active 5 h of the day, those with cardio-metabolic self-report weekly moderate to vigorous physical activity disease performed a lower intensity of activity compared (MVPA). Self-report disease status was used to define four to those with no disease. During waking hours, total time disease groups spanning cardio-metabolic health, which spent in each of the activity thresholds (light time, moderate included ‘Healthy’ (individuals who reported no disease), time and vigorous time) declined across cardio-metabolic ‘Cardiovascular disease (CVD)’, Type 2 diabetes minus disease groups. Both men and women with ‘Type 2 diabe- CVD, and ‘Type 2 diabetes + CVD’. tes + CVD’ performed half the level of MVPA when con- Between February 2013 and December 2015, a subset of sidered as MVPA1min or MVPA10min bouts. Inactive time individuals was invited to wear an Axivity AX3 wrist-worn was higher across all cardio-metabolic groups and a similar triaxial accelerometer for 7 consecutive days. Raw accelerom- pattern was observed for Inactivity30min, whereby those eter data were processed using GGIR V1.5-9 package (R Core with ‘Type 2 diabetes + CVD’ spent almost 7 h of the day in Team, Vienna, Austria) (https://cr an.rroject.or g/web/packag es/ 30 min inactivity bouts (Inactivity30min). GGIR/inde x .html) [3]. We defined MVPA using a 100 mil- Figure 1 shows the prospective associations of baseline ligravity (mg) cut-off, based on laboratory findings [4 ]. Simi- disease status with objective physical activity variables after larly, 6 metabolic equivalents (METs) is classified as vigorous 5 years of follow-up, when adjusting for confounders such as activity and was equivalent to an acceleration around 400 mg BMI. There were significant inverse associations of disease [4]. ‘Light’ (< 3 METs) was defined as anything between 40 and status with average acceleration across week and weekend 100 mg, and ‘Inactivity’ as anything below 40 mg [2]. Within days for both men and women. Similar patterns were seen each threshold, total activity time within waking hours was with bouted and unbouted MVPA, whereas disease status calculated (Light time, Moderate time, Vigorous time, Inactiv- demonstrated a slightly weaker positive association between ity time). Additionally, time spent in 1–5 min (MVPA1min) bouted and unbouted inactivity. Sensitivity analysis demon- and 10 min (MVPA10min) bouts of MVPA and time spent in strated similar associations. 30 min of inactivity (Inactivity30mins) was calculated. Discussion Statistical analysis In the largest objective cohort to date, there is a decrease in Due to the large sample size, any small difference in accel- objectively measured physical activity from healthy to CVD eration mean was significant, therefore these results are to diabetes patients which was consistent whether week or not reported. Multiple linear regression models were used weekend acceleration was measured, or bouted or unbouted to investigate the association between cardio-metabolic activity was used. These results demonstrate the usefulness disease and objective physical activity, after adjusting for of accelerometers in exploring the relationship between baseline covariates. Physical activity variables did not meet physical activity and chronic disease. The results also pro- assumptions of normality, therefore were transformed. All vide a platform for future novel explorations including the analyses were stratified by gender due to significant interac- temporal distribution and patterns of physical activity, sleep tion between gender and outcome variables. To determine and sedentary behaviour. the robustness of the results, sensitivity analysis was per- These devices make it possible to measure bouted and formed with a more stringent cut-point of 120 and 50 mg unbouted activity. We chose to focus on MVPA bouts in to define MVPA and inactivity, respectively. Analyses were 1 3 Acta Diabetologica Table 1 Physical activity acceleration values from Axivity in all participants (n = 52,424) according to disease status and stratified for gender Male (n = 24,880) Female (n = 27,544) Healthy CVD Type 2 Type 2 dia- Healthy CVD Type 2 Type 2 dia- (n = 11,232) (n = 11,996) diabetes betes + CVD (n = 14,960) (n = 11,746) diabetes betes + CVD minus CVD (n = 1218) minus CVD (n = 561) (n = 434) (n = 277) Age, years (SD) 54.3 (8.0) 59.6 (6.8) 59.1 (6.9) 61.0 (5.9) 53.6 (7.6) 58.5 (7.0) 58.6 (6.2) 59.9 (6.5) BMI, kg/m (SD) 26.3 (3.5) 28.4 (4.2) 29.5 (4.4) 31.4 (5.3) 25.2 (4.1) 28.0 (5.4) 31.4 (6.5) 33.2 (6.4) Physical activity  Average accel- eration values, mg (SD)   Daytime 42 (15) 36 (12) 34 (11) 31 (12) 44 (13) 38 (12) 35 (12) 31 (11) acceleration   Acceleration 0.63 (1.04) 0.69 (1.10) 0.69 (0.74) 0.79 (0.93) 0.54 (0.94) 0.57 (0.73) 0.70 (1.36) 0.72 (0.82) for least active 5 h   Acceleration 67 (28) 56 (22) 52 (18) 47 (25) 66 (23) 57 (18) 51 (18) 46 (17) for most active 5 h   Weekday 30 (6) 26 (8) 24 (7) 22 (11) 30 (8) 27 (8) 24 (7) 22 (7) acceleration across night and day   Weekend 30 (12) 25 (10) 23 (8) 21 (7) 30 (10) 26 (8) 23 (8) 21 (7) acceleration across night and day  Total time spent across differ - ent thresholds during waking time (min/ day)   Inactivity 588 (75) 604 (3) 615 (78) 624 (74) 568 (77) 583 (75) 599 (73) 617 (80) time   Light time 162 (47) 156 (47) 153 (50) 146 (48) 182 (46) 178 (48) 167 (55) 156 (53)   Moderate 96 (45) 79 (40) 72 (39) 61 (38) 104 (44) 87 (43) 75 (43) 62 (40) time   Vigorous time 6.12 (7.6) 3.58 (5.04) 2.64 (3.01) 1.87 (2.50) 4.8 (6.3) 2.8 (3.8) 2.1 (3.3) 1.5 (2.5)  Bouts of activ- ity during waking time (min/day)   MVPA10min 22 (28) 14 (20) 13 (19) 8 (19) 20 (25) 12 (19) 9 (17) 5 (12)   MVPA1min 23 (14) 18 (12) 16 (12) 13 (11) 25 (14) 20 (13) 16 (13) 13 (12)   Inactivity- 357 (124) 394 (128) 412 (134) 432 (133) 318 (115) 353 (122) 380 (139) 419 (141) 30min regression models, as it has been previously shown that bouts this reason, identifying associations with > 1 min bouts of of MVPA have stronger associations with metabolic health, MVPA is the most informative. Inactivity levels were high compared to unbouted activity [5]. Time spent in unbouted regardless of disease status, however in this study, associa- activity seems higher than what would be expected, with an tions between inactivity and cardio-metabolic were not as average of around 60 min/day for those with ‘Type 2 diabetes strong as other accelerometer variables. This most likely and CVD’, but this could capture sporadic arm movements reflects the methodological limitations in defining ‘inactiv - which cannot be separated from true physical activity. For ity’ using accelerometers. Due to the postural component, 1 3 Acta Diabetologica Fig. 1 Associations of disease group at baseline with objective were adjusted for age, BMI, Townsend Deprivation Index, ethnic- physical activity after an average 5 ± 1  years of follow-up. Weekday ity, smoking, fruit and vegetable intake, alcohol, self-report weekly acceleration and Inactivity30min were log transformed, and week- MVPA, follow-up time (Women = solid line + triangle, Men = dotted end acceleration and MVPA1min were log+1 transformed. Models line + square). T2D type 2 diabetes it is currently difficult to distinguish between light physical Compliance with ethical standards activity, e.g., standing, and sedentary behaviour (i.e., reclin- Conflict of interest MIT is a founder of Changing Health Ltd, a digital ing/sitting), but efforts are underway to validate and define education company. SC, HF, JC, MC, AB declare that they have no con- this behaviour using accelerometers, as well as sleep. An flict of interest. important limitation with this study is that the direction of causality cannot be distinguished. Human and animal rights All procedures followed were in accordance with the ethical standards of the responsible committee on human experi- Overall, strong and consistent relationships between mentation (institutional and national) and with the Helsinki Declaration cardio-metabolic disease and triaxial accelerometry, dem- of 1975, as revised in 2008 (5). onstrate enhanced measurement opportunities and greater insights going forward. Informed consent Informed consent was obtained from all patients for being included in the study. Acknowledgements This research was conducted using the UK Biobank resource. The authors would like to thank the UK Biobank Open Access This article is distributed under the terms of the Crea- participants and investigators for making this study possible. We would tive Commons Attribution 4.0 International License (http://creat iveco also like to acknowledge support from the Newcastle University Centre mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- for Ageing and Vitality sponsored by the BBSRC, EPSRC, ESRC and tion, and reproduction in any medium, provided you give appropriate MRC for providing support. credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Author contribution statement MIT and SC designed the study analy- sis. SC, HF, JYC, MC, AB, and MIT assisted in data analysis and References interpretation. SC and HF wrote the manuscript. All authors critically reviewed the manuscript and approved the final version for publication. ‘MIT is the guarantor of this work’. 1. Healy GN et al (2008) Objectively measured sedentary time, physi- cal activity, and metabolic risk: the Australian Diabetes, Obesity and Lifestyle Study (AusDiab). Diabetes Care 31(2):369–371 Funding MIT was supported by a Senior Research Fellowship from 2. Rowlands AV et al (2017) Accelerometer-assessed physical activity the National Institute for Health Research. in epidemiology. Med Sci Sport Exerc 4:1 1 3 Acta Diabetologica 3. Doherty A et al (2017) Large scale population assessment of physi- 5. Strath SJ, Holleman RG, Ronis DL, Swartz AM, Richardson CR cal activity using wrist worn accelerometers: the UK Biobank (2008) Objective physical activity accumulation in bouts and study. PLoS One 12(2):1–14. https://doi.org/10.1371/journal. nonbouts and relation to markers of obesity in US adults. Prev pone.0169649 Chronic Dis 5(4):A131 4. Hildebrand M, Van Hees VT, Hansen BH, Ekelund U (2014) Age group comparability of raw accelerometer output from wrist- and hip-worn monitors. Med Sci Sport Exerc 46(9):1816–1824 1 3

Journal

Acta DiabetologicaSpringer Journals

Published: May 28, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off