REVIEWING PUBLIC HEALTH INTERVENTIONS ON REDUCING SEDENTARY BEHAVIOR IN WORKPLACES

ABSTRACT


INTRODUCTION
Sedentary behavior, characterized by a low energy expenditure (1.5 metabolic equivalent or less), has been found to increase the risk of non-communicable diseases (NCDs) such as cardiovascular diseases, diabetes, obesity, cancer and premature mortality (Biswas et al., 2015;Thorp, Owen, Neuhaus, & Dunstan, 2011;Mark S Tremblay et al., 2017;Mark Stephen Tremblay, Colley, Saunders, Healy, & Owen, 2010).NCDs contribute 80% of the total global burden diseases and require immediate action to reduce the risks (Wang & Wang, 2020).Aligned with the global action plan for the prevention and control of NCDs, member states of WHO agreed that by 2025, the prevalence of insufficient physical activity must be reduced by 10% (WHO, 2013).To achieve that goal, public health interventions in reducing sedentary behavior have been massively developed, especially in workplace settings (Dunstan et al., 2013).This is due to the fact that sedentary behaviors are mostly performed by office-based workers requiring long prolonged sitting while working with less physical activity performed (Dunstan et al., 2013).Therefore, specific public health interventions aiming at reducing sedentary behavior in workplaces are highly needed, not only emanated to address ergonomic problems (i.e., musculoskeletal disorders) but also focused to reduce and interrupt sitting time.
Changing behavior is a complex and major challenge, it is difficult to sustain (Bouton, 2014).Thus, identifying effective public health interventions is important to do.This review has the main objective to identify and examine the most effective and appropriate practice to reduce sedentary behavior in the occupational environment.The main hypothesis of this review is "there is a significant reduction and interruption of sitting time among office-based workers after the implementation of the interventions".

METHOD
Database searching was conducted in PubMed using the following search strategy: ("Sedentary Behavior"[MAJR]) AND ("office" [tiab] OR "workplace" [tiab])) AND ("adults" [tiab] OR "workers") AND ("intervention*" [tiab] or "program").The search strategy resulted 124 relevant papers published within 2010-2021.Manual scanning and reviewing was done to deem eligibility of the studies as well as to eliminate irrelevant study design and topic.For example, studies which did not carry out public health intervention and studies only exploring the experiences of workers were excluded to the review group.In addition, to assess the quality of the articles, Cochrane's risk of bias for randomized control trial was used as a guideline.This method has identified four eligible articles to include in this review.Those four articles were deemed eligible as they included (i) workplace setting (ii) carried out public health intervention at reducing sedentary behavior and/or increasing physical activity (iii) measured sedentary behavior as a primary and/or secondary outcome.

Recruitment:
Participants were recruited from 35 departments at 4 worksites in Texas, USA from 2010 through 2013 Attrition: Among the 3 arms, drop-out was estimated at 42%

Type of intervention:
A

Length of follow-up:
6 months

Change in outcome:
Consistent attendees in the Booster Break arm showed a greater weekly pedometer counts and had decreased their sedentary time compared to other intervention groups.

Author's conclusion:
Booster Breaks intervention is potential to achieve significant positive results in reducing sedentary behavior and increasing physical activity.However, it requires sufficient fidelity, dose and routine practice in the workplace.

Change in outcome:
Sedentary time among office workers during working hours was significantly reduced and the light intensity activity was increased after implementing a participatory workplace intervention program.However, there was no clear evidence which one of the 3 interventions had a greater impact in improving occupational sedentary behavior.

Author's conclusion:
Participatory workplace intervention, especially office work modification (i.e., providing treadmill in the workstation) is potential to reduce sedentary behavior among office workers -Diary to record times in bed, work, sleep quality and daytime sleepiness -Self-reported physical activity and sedentary behavior; measured using validated questions (secondary outcome)

Length of follow-up:
6 months

Fidelity:
-Study protocol published -The 3 interventions had been completed by 170 out of 263 participants enrolled

Change in outcome:
No changes in the primary outcome

Author's conclusion:
Both multi-component interventions showed unsuccessful result at increasing physical activity and reducing sedentary time

Recruitment:
Participants were recruited from 3 departments of a university and 3 departments of an environmental agency in East Flanders.
Recruitment information was sent via phone and email.

Contact Intensity:
-On-screen-messages every 30 minutes (could ask additional sections if users were interested)

Intervention Content Intervention Evaluation
Theory: Theory-driven, web-based, computer-tailored advice (based on Theory of Planned Behavior [TPB])

Strategies/components:
1. Computer tailored intervention: providing personalized feedback on participants' sitting time after completing an assessment questionnaire; feedback messages will be available on the users' screen.An action plan offered 2. Generic Intervention: providing generic information about the importance of Study Design:

3-arm cluster randomized controlled trial
Primary outcome measure: Device-measured sedentary behavior (sitting time, standing time, breaks from sitting were measured using ActivPal) Other measures: -Self-reported sedentary behavior -Day log to record the type of day -Website usage statistics

Length of follow-up:
3 months

Fidelity:
-Pre-specified study protocol reducing and interrupting sitting

Intervention Outcomes
Change in outcome: -Self-reported sitting measurement: Tailored group showed greater reduction in their sitting time (at work and during leisure time) than those in the generic group Objective measurement: no changes in total sitting time between the 3 group; tailored group showed a slight increase in breaks at work

Author's conclusion:
A web-based computer-tailored intervention may have potential in reducing and interrupting sitting at work

Discussion
The aim of this literature review is to critically evaluate the implemented public health interventions at reducing sedentary behavior among office-based workers.Among the four studies included, objective measurement to detect sedentary behavior (including sitting, standing, and stepping) was done using a high validity accelerometry-based device either ActivPal or ActiGraph (An, Kim, & Lee, 2017;De Cocker, De Bourdeaudhuij, Cardon, & Vandelanotte, 2015), except the study conducted by Taylor et al. (2016).They were utilizing pedometers and validated questionnaires to measure physical activity and sedentary time while self-reported checklist (potential recall and social desirability biases) was used to log sedentary leisure time.This study was aligned with the study conducted by Parry et al. (2013) in terms of their primary aim, which was increasing physical activity while working.Conversely, the other study more focused on reducing sitting time (De Cocker, De Bourdeaudhuij, Cardon, & Vandelanotte, 2016) while Nooijen et al. (2020) assessed both increased physical activity and reduced sedentary behavior using multicomponent interventions.Therefore, due to the nature of different focuses of the studies, it is noteworthy that different measurement instruments used in this study have created difficulties in making comparative analyses.
The blinding of participants and personnel of these studies were not feasible; only Nooijen et al. (2020) performed blinding of personnel involved in data collection and processing, while the researcher was not blinded to group allocation in the study by Parry et al., (2013).The other two did not specify participant and personnel blinding.Most of them addressed this issue by adhering to their pre-specified intervention components, while only Taylor et al. (2016) failed to do so.The fidelity of the Booster Break program was not clearly outlined which made the effectiveness of this intervention limited.
Furthermore, all studies included in this review were using a randomized controlled trial as their study design.Risks of contamination between groups were reduced given the cluster randomization methods were employed in these studies except for Parry et al. (2013) study.However, Parry et al. (2013) did not use individuals and/or team departments in the same organization as the unit of randomization, hence, it reduced the opportunity of cross contamination risks.Nevertheless, although a cluster randomized was a strength in the study by Nooijen et al. (2020), risks of cross contamination could not be ruled out due to the randomization of office workers into the same teams.
Of the four studies reviewed, the attrition rates were relatively high.However, Parry et al. (2013) and Taylor et al. (2016) did not clearly explain reasons for attrition rate.Although only based on researchers' speculations, Nooijen et al. (2020) with an overall 53.38% attrition rate outlined the reasons behind the number.They speculated that high workload, high staff turnover, and lack of time in doing physical activity were the reasons for participants not completing the intervention.Likewise, De Cocker et al. (2016) explained the drop-out rate based on two different points of views (methodological and compliance).ActivPal, a monitoring device, was not randomly allocated as it was better to provide the monitoring device to those who agreed to wear it based on the prescheduled time.This had lowered the drop-out rate compared to those who did not wear ActivPal and only completed the questionnaires.Additionally, all studies employed the analysis principle of intention to treat (ITT) to reduce the potential bias arising from missing values with detailed imputation steps outlined in the study by Nooijen et al. (2020).
Moreover, the external validity of the studies is another point to consider.82% of the study sample population in both Taylor et al. (2016) and Parry et al. (2013) were women.This potentially could create selection bias and reduce the generalizability which warrants further investigations.The generalizability limitation was further noticed in the study by Parry et al. (2013) due to the recruitment of the study participants; they did the recruitment and delivered the intervention within the same organization.Instead, the other three studies recruited the participants from different companies (De Cocker et al., 2016;Nooijen et al., 2020;Taylor et al., 2016).The findings applicability to the general population were also limited, as the three studies reviewed did not include any explanations on ethnicities and only Taylor et al. (2016) reported the diversity of the study participants; racially and ethnically diverse (35% African American, 33% non-Hispanic white, and 32% Hispanic).
Overall, the findings from four different studies included in this review showed conflicting outcomes.This may be due to the nature of their approaches used during delivering the interventions (single-component intervention vs. multi-component strategies).On one hand, surprisingly, from the three multi-component intervention strategies conducted, Nooijen et al. (2020) showed that this strategy was unsuccessful and was not effective; no changes in the primary outcome of the study (sedentary behavior was not reduced and physical activity was not increased).On the other hand, a single-component intervention targeting environmental level reported a promising result in reducing sedentary behavior among office workers by modifying workstation (i.e., providing treadmill) (Parry et al., 2013).Although the availability of physical activity equipment such as treadmills in the workstation has conferred benefits in increasing physical activity, this intervention did not confirm the hypothesis of reduced and/or interrupted sedentary behavior (prolonged sitting) among office-based workers.

CONCLUSION
There have been an increasing number of studies addressing sedentary behavior with successful outcomes, however, there is still a large number of potential gaps found, especially the long-term effectiveness of different promising interventions (Chu et al., 2016).Future studies also need to focus on identifying which population groups that will receive the most benefit from the interventions implemented, this highlights the urgent need to evaluate health inequities that exist among different cultures and ethnicities (Newman, Baum, Javanparast, O'Rourke, & Carlon, 2015).Thus, future intervention implementation will be more appropriate and friendly towards the general population.
World Health Organization.(2013).Global action plan for the prevention and control of noncommunicable diseases 2013-2020: World Health Organization