TY - JOUR

T1 - Accelerometer-Measured Physical Activity at Work and Need for Recovery

T2 - A Compositional Analysis of Cross-sectional Data

AU - Stevens, Matthew L.

AU - Crowley, Patrick

AU - Rasmussen, Charlotte L.

AU - Hallman, David M.

AU - Mortensen, Ole S.

AU - Nygård, Clas Håkan

AU - Holtermann, Andreas

PY - 2020

Y1 - 2020

N2 - Previous research has shown strong associations between occupational physical activity (OPA) and need for recovery (NFR). However this research has only utilized self-reported measures of OPA which may be biased. Thus, there is a need for investigating if the previously documented association between self-reported OPA and NFR can be found when using technical measures of OPA. There is also the need to investigate whether older workers are particularly susceptible to increased NFR, since age-related declines in physical capacity mean that it is likely these workers will have a higher NFR for a given physical activity. The aim of this study was to investigate the association between technically measured OPA and NFR, and whether this relationship is modified by age. Methods: This study utilized data from the Danish Physical Activity Cohort with Objective Measurements cohort - comprising Danish workers (n = 840) from the cleaning, manufacturing, and transportation sectors. OPA was measured by accelerometers attached to the thigh and upper back for at least one work day and classified into four physical behaviour categories (sedentary, standing, light, or moderate/vigorous). NFR was measured using a shortened version of the Danish NFR scale. Analysis was conducted using linear regression and isotemporal substitution analyses for compositional data. Results: The overall association between OPA and NFR was statistically significant in the unadjusted model (P < 0.001), but not when adjusted for age, sex, occupation, and shift work (P = 0.166). Isotemporal substitution showed small but significant reductions in NFR when increasing sedentary time relative to other behaviours (adjusted: ΔNFR = -0.010 [-0.019; -0.001]). There were no significant interactions between age and OPA (P = 0.409). Conclusions: This study found significant associations between OPA and NFR, but the effect sizes were small. Reallocating 30 min to sedentary behaviours from other behaviours was associated with a reduced NFR, but the effect size may not be practically relevant. Moreover, no clear modifying effects of age were identified.

AB - Previous research has shown strong associations between occupational physical activity (OPA) and need for recovery (NFR). However this research has only utilized self-reported measures of OPA which may be biased. Thus, there is a need for investigating if the previously documented association between self-reported OPA and NFR can be found when using technical measures of OPA. There is also the need to investigate whether older workers are particularly susceptible to increased NFR, since age-related declines in physical capacity mean that it is likely these workers will have a higher NFR for a given physical activity. The aim of this study was to investigate the association between technically measured OPA and NFR, and whether this relationship is modified by age. Methods: This study utilized data from the Danish Physical Activity Cohort with Objective Measurements cohort - comprising Danish workers (n = 840) from the cleaning, manufacturing, and transportation sectors. OPA was measured by accelerometers attached to the thigh and upper back for at least one work day and classified into four physical behaviour categories (sedentary, standing, light, or moderate/vigorous). NFR was measured using a shortened version of the Danish NFR scale. Analysis was conducted using linear regression and isotemporal substitution analyses for compositional data. Results: The overall association between OPA and NFR was statistically significant in the unadjusted model (P < 0.001), but not when adjusted for age, sex, occupation, and shift work (P = 0.166). Isotemporal substitution showed small but significant reductions in NFR when increasing sedentary time relative to other behaviours (adjusted: ΔNFR = -0.010 [-0.019; -0.001]). There were no significant interactions between age and OPA (P = 0.409). Conclusions: This study found significant associations between OPA and NFR, but the effect sizes were small. Reallocating 30 min to sedentary behaviours from other behaviours was associated with a reduced NFR, but the effect size may not be practically relevant. Moreover, no clear modifying effects of age were identified.

KW - blue-collar workers

KW - compositional data analysis

KW - need for recovery

KW - physical activity

KW - physical behaviour

KW - triaxial accelerometers

U2 - 10.1093/annweh/wxz095

DO - 10.1093/annweh/wxz095

M3 - Journal article

C2 - 31879769

AN - SCOPUS:85080846293

VL - 64

SP - 138

EP - 151

JO - Annals of Occupational Hygiene

JF - Annals of Occupational Hygiene

SN - 2398-7308

IS - 2

ER -