Standing desks for sedentary occupations: Assessing changes in satisfaction and health outcomes after six months of use

1 Introduction

Sedentary behaviors, or engaging in activities of low energy expenditure for prolonged periods, have been associated with adverse health outcomes indicated by various cardio-metabolic biomarkers and cardiovascular disease-associated and all-cause mortality [1]. Importantly, increased risk of adverse health outcomes is independent of whether individuals adhere to public-health recommendations for regular physical exercise [2]. Sedentary behavior can be quantitatively defined as any waking activity characterized by energy expenditures of less than 1.5 metabolic equivalents (for comparison, moderate to vigorous physical exercise expends 3–8 metabolic equivalents) [3]. When an individual engages in activities of musculoskeletal tension (e.g. running, walking, even standing upright), metabolic energy is proportionally expended [4].

Electronic monitoring devices, such as accelerometers, have helped shed light on the characteristics of sedentary behaviors in the context of modern daily life. It has been reported that the average adult engages in eight to nine hours of sedentary behavior daily, with estimates suggesting that one half to two-thirds of that occur at work [5, 6]. A small Australian study of office workers further discovered that sedentary behavior is significantly more sustained (bouts >30 min/less standing breaks) during work hours compared to non-work hours [7]. Interestingly, sustained sedentary behavior, independent of total sedentary time, has been negatively associated with metabolic risk factors including waist circumference (WC), body mass index (BMI), triglycerides, and 2 hour plasma glucose [8].

Individuals who reduce sedentary behavior have demonstrated improved health outcomes including: reductions in hip and waist circumference, body weight, body fat, low-density lipoprotein (LDL), total cholesterol as well as increases in high-density lipoprotein (HDL) [9, 10]. Even taking breaks during prolonged bouts of sedentary behavior has been independently associated with improvements in various obesity metrics [8, 11]. These findings culminated in recommendations by a panel of health experts in the United Kingdom that office workers should aim to spend 2–4 hours per day on their feet [12]. Toward that aim, a range of workplace interventions including participatory physical activity “challenges”, educational/awareness campaigns, self-monitoring strategies, and workplace furnishing interventions have been deployed [12].

Adjustable height workstations, or standing desks, have emerged as one of the most readily available and accepted methods for promoting a more active working environment [13, 14]. Customization of the workstation allows the employee to adjust the height of the computer and keyboard, offering the freedom to perform work tasks from a seated to standing position at-will and with relative ease. Despite the promising findings of standing desk implementations, there is limited information regarding the long-term health outcomes of standing desk interventions. Additionally, the utilization of standing desks within a hospital setting is not well-studied. Hospital staff, in various capacities, are a critical population not only because of their relatively poor health profile, but also because of their visibility as lifestyle role models in the clinical setting [15–17]. The purpose of the present investigation was to explore the impact of a six month standing desk intervention on overweight hospital employees who self-reported sedentary behaviors of six or more hours per workday.

2 Materials and methods

All research complied with the ethical rules for human experimentation as stated in the Declaration of Helsinki. This study was reviewed and approved by the Institutional Review Board and all participants completed informed consent prior to participation in the study. Inclusion criteria were: 18 years of age or older, self-reported sitting at their desk six or more hours per workday, BMI of 25 kg/m² or higher, and having a personal workstation and supervisor permission to participate/allow the installation of the standing desk. Participants were recruited from a large military medical facility in the Pacific. Using the facilities SharePoint 360 announcement and news section, an announcement for the research study was made along with the posting of fliers across the facility.

Target enrollment was 30 participants, 10 in the control group and 20 in the intervention group. We assigned more participants to the intervention group to maximize sample size for assessment of the effect of intervention. A total of 35 participants were enrolled (11 in the control group and 24 in the intervention group) to account for potential dropouts. Blocked Randomization, the process of assigning participants to the control or intervention group, was carried out in blocks of 10 participants for the first 20 participants. The last 15 participants were all assigned to the intervention group. For the first two blocks, group assignments were made based on random assignment to either the control or intervention group, with 10 participants assigned to the control group and 10 to the intervention group. Overall, a total of 35 participants (7 males and 28 females) were enrolled though nine failed to complete the study. The remaining 18 intervention and 8 control participants were used for analysis. Reasons for not completing the study included two participants who immediately declined participation when they were randomized to the control group, one who became pregnant and was excluded from analysis, three lost to follow-up, and three participants who stopped working at the facility during the study.

Height, weight, and waist circumference were collected by asking the participants to remove their shoes and anything in their pockets. A calibrated, medical grade digital scale was used to record participants’ weight. Utilizing a height rod, height was measured and rounded up to the nearest centimeter. Waist circumference was measured utilizing a soft retractable tape measure. Waist circumference measurements were performed at the natural waist circumference measured against the skin about halfway between the naval and lower end of the sternum. The tape measure was kept parallel to the floor and the measurement was completed three times, rounding down the average of the three measurements to the nearest centimeter. With few exceptions, all measurements from participants were completed by the same research coordinator. Body Mass Index was calculated by dividing the weight in kilograms by height in meters squared.

Once enrolled, all participants had baseline height, weight, and waist circumference measurements completed. Participants in the intervention group additionally completed a workplace satisfaction survey and a three-day diet recall form during initial study enrollment. The following week, an adjustable height desktop accessory (Kangaroo© standing desk by Ergo Desktop, Celina OH; main work surface 28^″×24^″, adjus16.6^″ above desk; monitor shelf 16^″×9.25^″, adjustable 6.5^″ above work surface) was installed within the respective workplace of each intervention participant. During installation, intervention participants were provided with an anti-fatigue mat and were instructed on the proper use and adjustment of the standing desk. Participants with standing desks were informed of numerous free resources available to assist with tracking of calorie expenditure and desk usage. Participants were also encouraged to gradually increase the standing desk configuration from a starting target of 10 min/hour, to minimize any short-term discomfort that may arise from musculoskeletal adjustments during the initial stages of transition [18]. Participants were not given any follow-up reminders or specific targets for frequency or length of utilization of the standing desk. Participants were provided the contact information of the research team and encouraged to contact them for any problems or questions that may arise. Female participants were instructed to inform the research team in the event of becoming pregnant and to obtain a letter of physician’s support if they chose to continue participation in the study thereafter.

Participants were administered a pre-and post- installation survey evaluating personal health/physical activity goals and standing desk satisfaction. Three months after installation, participants in the intervention group were provided a desk-use log to record daily sitting/standing times throughout an eight hour workday for five days. That same week, they were fitted with activPAL 3 micro (PAL technologies Ltd., Scotland, UK), a uniaxial device demonstrated to accurately differentiate between sitting/lying, standing and stepping in adults [19]. Prior to recording, all devices were initialized on a Windows compatible PC according to manufacturer’s instructions. The device was worn on the thigh of the dominant side of the participant and secured using a water-proof nitrile sleeve (3 M Tegaderm, Location, St. Paul, MN) for one week. Finally, at the end of the six-month period, the workplace satisfaction survey, height, weight, and waist circumference were repeated for all participants.

Paired t-tests were used to assess changes over time in BMI and waist circumference for both the control and intervention groups. Analysis of covariance was done to evaluate group differences in BMI and waist circumference at 6 months, adjusted for baseline levels at time of enrollment. Unpaired t-tests were used to assess group differences in mean BMI and waist circumference at the start of study. Wilcoxon signed rank tests were used to compare changes in workplace satisfaction between baseline and 6 months. Bland-Altman plots were used to represent deviation between subjective and objective measures of physical activity during the workday. Intra-class correlation coefficients (ICC) were calculated (generally rated as poor (0–0.39), moderate (0.4–0.74), and excellent (0.75–1). Paired t-test was used to evaluate differences between self-reported and actual standing time. A significance level of 0.05 was used for all analyses. All analyses were conducted using SAS statistical software version 9.4 (SAS Institute, Cary, NC).

3 Results

Table 1 summarizes mean BMI and waist circumference levels at baseline and six months later for the control and intervention groups. Even though participants were randomized, participants assigned to the control group had significantly higher BMI and waist circumference at baseline compared to those in the intervention group (35.3 vs. 30.8 for BMI, p = 0.037; and 41.2 in vs. 36.7 in for waist circumference, p = 0.027). For both groups, mean BMI and waist circumference did not change significantly over the six months of the study (mean change in BMI = –0.08, p = 0.694, for intervention group, and –0.06, p = 0.940, for control group; mean change in waist circumference = 0.3 inches, p = 0.461, for intervention group, and –0.01 inches, p = 0.986, for controls). Analysis of covariance did not find a significant group difference for mean BMI and waist circumference at six months, adjusted for baseline levels (p = 0.729 and p = 0.856, respectively).

One of the criteria for enrollment in the study was sitting for a minimum of six hours (360 minutes) during an average eight-hour workday. Thus, at baseline, it can be deduced that participants engaged in non-sedentary workplace behavior for no more than two hours (120 minutes) or 25% of the work day. At the three-month intervention point, the self-reported average time engaged in non-sedentary behavior was 254 minutes (±126), thus, total time engaged in sedentary workplace behavior could be estimated at about 226 minutes (3.76 hours) per workday, a decrease of about 134 minutes (2.24 hours) or 25% from baseline. Only one intervention participant reported <120 minutes of non-sedentary behavior three months after the installation of the standing desk, while non-sedentary times reported by all other participants ranged from 129 to 586 minutes. Three month changes in sedentary/non-sedentary workplace behavior based on baseline enrollment criteria are displayed in Fig. 1.

OPEN IN VIEWER

Fig.1

Three month change in sedentary/non-sedentary workplace behavior for the intervention group. ^*Baseline averages calculated based on study inclusion criteria of minimum 6 hour sedentary time during 8 hour workday (120 min standing maximum).

To address the reliability of self-reported estimates of daily sitting/standing time at the three month intervention period, we performed a comparison of self-reported daily averages and daily averages recorded by accelerometer for one work week. By this method, we calculated that self-reported daily standing minutes and actual daily standing minutes was 52.9% of the workday versus 42.7% (based on 5 day average). This translated to about a 10% overestimation of daily standing time or the equivalent of 50 minutes, but this discrepancy was not significant (p > 0.05). This indicates the reliability of the daily log of workday position as a measure of sedentary/non-sedentary time during the intervention trial (Table 2).

Another notable study finding was the effect of the implementation of standing desks on personal workplace satisfaction. The number of respondents who indicated they were “extremely satisfied” with the comfort of their office furnishings, ability to adjust furniture to meet needs, and overall satisfaction with personal workplace” increased dramatically, from 6% before the installation of the standing desk to over 67% at the six-month post installation mark. Additionally, the percentage of respondents who reported being either “somewhat” or “extremely satisfied” with the comfort of their office furnishings (34%), ability to adjust furniture to meet their needs (72%), and personal workplace satisfaction (44%) also demonstrated an increase. Overall feedback was positive, with only one case of use-related soreness/fatigue reported. The most common obstacle for usage was the fact that most participants used two monitors but the standing desk was equipped to hold only one monitor.

Due to the growing popularity of standing desks and other behavioral interventions in reducing sedentary workplace time, we sought to explore the long-term impact of a standing desk intervention in a population of overweight hospital employees. The self-reported usage of the standing desk after three months was four hours. This was a substantial amount of time given that participants reported sitting in excess of six hours per workday (or 75% of an 8-hour workday) prior to the installation of a standing desk. These findings corroborate earlier reports describing a two hour and 21% reduction in workplace sedentary time in comparable Australian and American studies, respectively [20, 21]. Caloric estimates for standing versus sitting vary in the literature from +0.83 calories per minute to+1.36 calories a minute [22, 23]. Combining these estimates with the average reported standing time reported at the three month time point suggests participants possibly increased their caloric expenditure between 203 and 332 calories per day with this simple behavioral change, all else being equal. Extending the estimated increase in caloric expenditure over the course of six months of use, assuming no other changes in behavior, it was expected that some weight-associated metrics would be improved among the intervention group.

In this study, standing desk use had no significant impact on weight lost, BMI, or WC after a six-month intervention. This is in contrast not only to the expected manifestation of increased daily caloric expenditure, but also to reports correlating sedentary office behavior to increased (proportional) risk of negative weight-related outcomes [24]. Our work, however, supports the findings of complementary studies which reported similar decreases in sedentary work (and even non-work) behavior in the absence of changes in fat mass, BMI, or WC after a three month intervention [20]. Even a standing desk study reporting a 50% reduction in office sedentary behavior (double the reduction achieved herein) found no significant differences across various markers of cardio metabolic risk [25]. Interestingly, some standing desk studies have been able to track intervention-related improvements across a variety of health outcomes including blood pressure and pulse wave velocity in overweight to pre-diabetic individuals, though these results were modest and acute in nature [26].

A possible reason for the absence of positive health outcomes in our study may lie in the limited ability to control for a variety of potentially confounding factors. For example, while we performed a modest assessment (3-day diet recall) of participant’s diets prior to installation of the standing desk and three months after, continuous monitoring of participants’ diets throughout the course of the study was not feasible. However, from our limited data, a dietician did not identify significant changes in self-reported nutritional habits over the course of the intervention period. Along similar lines, a recent study has reported that breaks of light physical activity during prolonged sitting periods does not impact appetite [27], though the acute nature of the study may not hold true in long-term interventions. The idea of “compensatory” responses to sedentary interventions has been around for some time, particularly where physical activity is concerned. Several investigators have put forth the idea that the non-intervention period may be susceptible to alterations as a response to perceived “tiredness” or “fatigue” from sedentary intervention. On the other hand, Rowland provides some evidence which supports the fact that exercise participation does not inadvertently impact sedentary behavior during non-exercise (non-intervention) periods [28].

Finally, the lack of detectable improvements in long-term health outcomes may be due to the subjectivity of self-reported sedentary behavior (e.g. overestimating standing time/underestimating sitting time) and/or limitations of the selected health outcome measures. While some studies have reported high concordance between self-reported and objectively measured sitting time [29, 30], a 2016 study of 236 desk-based government employees recruited across several agencies called into question the reliability of self-reported metrics to accurately reflect incidental physical activity during work hours [31]. Our monitoring of activity suggests that self-report was 10% higher than actual usage. It is possible that long term health indicators require a heightened cardiometabolic sensitivity. For example, Mun et al 2017 points out that studies using body fat percentage as a metric are more consistent in their ability to demonstrate the relationship between sedentary behavior and obesity compared to studies using surrogate measures (BMI, WC, etc.) [32].

The study herein suggests that the health benefits of long-term decreases in sedentary behavior may be masked by confounding factors or underestimated by indirect measures of obesity/cardio-metabolic health. Despite that, it was clear that participants were much more satisfied with their personal workspace after the installation of a standing desk (58% increase). Along the lines of personal workplace satisfaction, many standing desk studies have similarly reported improvements in non-physiological factors including comfort, mood, and overall well-being [33]. Feedback from our study’s participants appear to reflect a similar pattern, with largely positive remarks including “The standing desk helped relieve pressure on my legs I had been experiencing from prolonged seating” and “increase energy (sic), less tiredness”. The largest complaint from participants in this study was that the surface area of the raised platform was often not large enough to accommodate two monitors comfortably, a caveat shared by a similar standing desk accessories [18]. Though traditional sit to stand desks alleviate the limitation of available workspace in the more affordable, standing desk attachments, the overall reductions in sedentary time obtained by the latter average 83 minutes per workday [21, 34], similar to the average 100 minute reduction concluded in a recent review of randomized controlled trials using (mostly) traditional adjustable height workstations [35].

While the long-term health benefits of standing workstations remain to be substantiated in future work, their effectiveness in reducing sedentary workplace behavior continues to be demonstrated across various paradigms and populations. Moreover, standing desks continue to be associated with positive wellness indicators including increased moderate to vigorous physical activity during the workweek, [36] reductions in perceived discomfort, and increased productivity [37]. Given the relatively low cost nature, minimal footprint, and improvements in user satisfaction observed in this study, it is worthwhile to continue to investigate the long-term health effects of reduced sedentary workplace behavior and to optimize workstation design and institutional policy to achieve lasting reductions in sedentary behavior.

5 Conclusions

Excessive sedentary behavior is linked to adverse health effects. As the majority of sedentary behavior occurs during working hours, changing the workplace environment is one logical opportunity to encourage employees to decrease their time in a seated position. Our findings indicate that standing desks indeed represent a low cost and effective option for employees to reduce their sedentary behavior. Though the use of a standing desk for six months in an overweight group of administrative hospital employees did not produce expected changes in long term health outcomes, employees reported significant improvements in overall workplace satisfaction. Future work is required to more sensitively gauge changes in long term health outcomes. This study also demonstrated a significant reduction in sedentary behavior regardless of subjective or objective reporting methods. Due to the feasibility of implementation, the positive user feedback and the significant reduction in workplace sedentary behavior observe, we advocate further investigation of the effect of standing desks on the long term health outcomes of potential users, particularly in at-risk populations.

Conflict of interest

None to report.