Correction of bias in self-reported sitting time among office workers – a study based on compositional data analysis

Abstract

Correction of bias in self-reported sitting time among office workers -a study based on compositional data analysis by Coenen P, Mathiassen SE, van der Beek AJ, Hallman DM Being the first to develop a calibration model using compositional data analysis, we found that a model in which "true" occupational sitting was estimated from self-reported sitting, or even additional predictors, led to substantially more correct estimates than if only self-reports were used. Our models can be used for post-hoc improvement of self-reported sitting data and while designing future studies.Coenen P, Mathiassen SE, van der Beek AJ, Hallman DM. Correction of bias in self-reported sitting time among office workers -a study based on compositional data analysis.Objective Emerging evidence suggests that excessive sitting has negative health effects. However, this evidence largely relies on research using self-reported sitting time, which is known to be biased. To correct this bias, we aimed at developing a calibration model estimating "true" sitting from self-reported sitting.Methods Occupational sitting time was estimated by self-reports (the International Physical Activity Questionnaire) and objective measurements (thigh-worn accelerometer) among 99 Swedish office workers at a governmental agency, at baseline and 3 and 12 months afterwards. Following compositional data analysis procedures, both sitting estimates were transformed into isometric log-ratios (ILR). This effectively addresses that times spent in various activities are inherently dependent and can be presented as values of only 0−100%. Linear regression was used to develop a simple calibration model estimating objectively measured "true" sitting ILR (dependent variable) from self-reported sitting ILR (independent variable). Additional self-reported variables were then added to construct a full calibration model. Performance of the models was assessed by root-meansquare (RMS) differences between estimated and objectively measured values. Models developed on baseline data were validated using the follow-up datasets.Results Uncalibrated self-reported sitting ILR showed an RMS error of 0.767. Simple and full calibration models (incorporating body mass index, office type, and gender) reduced this error to 0.422 (55%) and 0.398 (52%), respectively. In the validations, model performance decreased to 57%/62% (simple models) and 57%/62% (full models) for the two follow-up data sets, respectively.Conclusion Calibration adjusting for errors in self-reported sitting led to substantially more correct estimates of "true" sitting than uncalibrated self-reports. Validation indicated that model performance would change somewhat in new datasets and that full models perform no better than simple models, but calibration remained effective.

Time-Based Data in Occupational Studies: The Whys, the Hows, and Some Remaining Challenges in Compositional Data Analysis (CoDA)

“…nutritional epidemiology ( Leite, 2016 ), geology ( Tolosana-Delgado and von Eynatten, 2009 ), and chemistry ( Buccianti and Pawlowsky-Glahn, 2005 ). However, in public and occupational health sciences, CoDA has gained attention only recently ( Pedišić, 2014 ; Chastin et al , 2015 ; Pedisic et al , 2017 ; Dumuid et al , 2018b ; Foley et al , 2018 ; Bauman et al , 2019 ), with few papers devoted to exposures at work ( Gupta et al , 2018a , 2019 ; Rasmussen et al , 2018 ; Hallman et al , 2019 ; Coenen et al , 2020 ).…”

“…too little non-sitting ( van der Ploeg and Hillsdon, 2017 ; Stamatakis et al , 2019 ). Analyzing compositional data using standard methods may even in some cases lead to absurd results, such as confidence intervals (CIs) including values less than 0% or more than 100% time ( Coenen et al , 2020 ).…”

“…Different principles for constructing ratios and performing the log-transformation have been described in the literature ( Pawlowsky-Glahn et al , 2015 ). In current research, the most widely used is the isometric log-ratio (ilr) transformation ( Carson et al , 2016 ; Biddle et al , 2018 ; Dumuid et al , 2018a , b ; Foley et al , 2018 ; Hallman et al , 2019 ; Coenen et al , 2020 ). Using the ilr transformation on a composition with d parts results in d − 1 orthonormal ilr coordinates.…”

Calibration of Self-Reported Time Spent Sitting, Standing and Walking among Office Workers: A Compositional Data Analysis

“…A novelty of this study is the use of the complete composition of multiple self-reported physical behaviors to predict ‘true’ time use. In a previous calibration study in the same population, we focused only on sitting [35] and expressed sitting time using a single ILR (i.e., sit/nonsit) without considering that self-reported information on behaviors occurring during non-sitting (i.e., standing and walking) might be of additional predictive value for sitting. In that study, we found that the simple calibration model reduced RMS error for the sit/nonsit ILR to 55% of the error before calibration.…”

“…We have recently shown that calibration within a CoDA framework can be effective in improving the accuracy of self-reported time in a single behavior (sitting) among office workers [35]. It is likely that calibration of a particular behavior will be even more effective, that is, lead to more accurate estimates of ‘true’ behavior, if self-reports of all other component behaviors are also included as predictors.…”