It is well established that too much sedentary time is a health concern as it has been linked to numerous adverse health outcomes, such as increased risk of cardiovascular disease (Jingjie et al., 2022), Type II diabetes (van der Berg et al., 2016), and all-cause mortality (Patterson et al., 2018). Approximately 30% of a person’s waking hours is spent standing, making this postural behavior an important area of study (Hamer & Stamatakis, 2020). While standing has been promoted to reduce sedentary time (Hwang et al., 2022), whether more time spent standing translates to measurable health benefits is mixed. Possibly, the divergent findings in the literature may be a result of not all types of standing being the same. A recent paper published in the Journal for the Measurement of Physical Behaviour investigated whether active standing can be identified separately from passive standing using thigh-worn accelerometry (Kowalsky et al., 2024). The study involved 10 participants who wore an activPAL monitor on their thigh and stood for three 5-min periods in different conditions: passive standing (no movement), active standing (structured micromovements), and free-standing (participant’s choice of movement). The study developed and validated a machine learning model to classify active and passive standing. Kowalsky et al. (2024) compared the average acceleration values in 15-s epochs across the three conditions and distinguished that there is a detectable difference between active and standing behaviors. Then, a machine learning model was developed with leave-one-subject-out validation, which achieved 94% accuracy for classifying passive versus active standing using 5-s epochs. When applied to the free-standing data, the model showed an absolute average difference of 4.8% when compared to observation and an area under the curve of 0.71, indicating fair performance of the model. This study demonstrates the potential for using accelerometry and machine learning to broadly distinguish between different types of standing behaviors and encourages additional model refinement. This research is an important addition to the field and will stimulate further work in this area to better understand standing.
Advancements in the Measurement of Active Versus Passive Standing
Kowalsky et al. (2024) advanced the measurement of active versus passive standing from Anderson et al. (2019). While the additional movements and conditions utilized in Kowalsky’s protocol compared to Anderson et al. (2019) are a strength, we would like to further emphasize the usefulness of considering the three-dimensional aspect of movement and movement variability as inputs into the decision tree model. Despite both studies yielding similar accuracy in their tested controlled movements (Anderson et al., 2019; Kowalsky et al., 2024), the theoretical basis and likelihood of movements in the anterior–posterior and superior–inferior planes being highly applicable to differentiating types of standing are strong. Certainly, predictive modeling is dependent upon the protocol used to develop the algorithm. Model development/testing based on longer periods of free-living data using high-resolution cameras (e.g., chest-mounted cameras to serve as a criterion; Bach et al., 2022) may further advance the work of Kowalsky et al. (2024) and Anderson et al. (2019). The shared data set and Python code by Kowalsky et al. (2024) may serve as a useful starting point for the development of future machine learning algorithms to distinguish standing types based on full-day and/or weeklong monitor wear times.
The authors emphasize that distinguishing standing types from alternative device locations (i.e., wrist) may be fruitful due to the popularity of commercial monitors, but they position that devices worn on the thigh are likely the best (Kowalsky et al., 2024). We do not disagree but also want to highlight the potential utility of multiple monitor configurations that could be more accurate than a single device. For example, activPALs have been placed on the thigh as well as the sternum (Taraldsen et al., 2011), shin (Martin et al., 2015), and torso (Basset et al., 2014) to better distinguish sedentary postures. Presumably, a monitor on the thigh and shin (or possibly a third monitor on the torso; Wu et al., 2024) could be particularly useful for model developments to distinguish active versus passive standing, albeit at the cost of more research materials, additional computational complexity, increased participant burden, and possible compliance challenges. Future development to single-thigh-based active versus passive standing is warranted regarding the ideal model parameters, postures/movements tested during validation, and accelerometer settings (e.g., epochs, sampling rates). However, the consideration of multiple monitor configurations and their average acceleration versus acceleration variability outcomes in three-dimensional planes is likely important for future developments. In general, the authors conservatively position the findings, emphasizing more model refinement because of the observed algorithms area under the curve although low absolute measurement error was observed. The stronger theoretical rationale for the model that relies on both the average and the variability of accelerometer thresholds across multiple planes is important. This, and the sharing of their data set/coding, encourages more methodological developments and better positions the trajectory for future measurement research targeting standing.
Physiological Impacts of Active Versus Passive Standing
The use of accelerometry and the novel machine learning model presented by Kowalsky et al. (2024) could be applicable in teasing out the physiological differences to stationary behaviors. Such behaviors are often grouped into one collective measure, but active standing may result in greater metabolic activity than passive standing. With active standing, the musculoskeletal system is more engaged than passive standing as there are frequent small movements (Kowalsky et al., 2024). Active standing may be superior to passive standing when stepping is not feasible, as it requires greater energy expenditure (Levine et al., 2000) and therefore may translate to greater metabolic health benefits. Consistent with this notion, marginal differences in energy expenditure are typically observed between passive standing and sitting (Betts et al., 2019). Furthermore, prolonged standing has been reported to be associated with low back pain (Andersen et al., 2007) and fatigue (Drury et al., 2008). Using an assembly task, passive standing exhibited worse fatigue rates in the lower legs and back compared with dynamic standing (Balasubramanian et al., 2009), suggesting that active standing may yield better musculoskeletal benefits than passive standing.
During passive standing, the gravity-induced hydrostatic stress and lack of a skeletal muscle pump result in blood pooling in the lower limbs. This pooling results in hemodynamic (e.g., decreased venous return), autonomic dysfunction (e.g., baroreceptor unloading), and metabolic (e.g., decreased glucose metabolism), that ultimately increases arterial stiffness and myocardial burden (Stoner et al., 2021). By involving the skeletal muscle pump, active standing may negate or attenuate some of these adverse cardiometabolic effects and yield positive cardiovascular effects. Studies examining the cardiometabolic benefits of active versus passive standing using free-living data are warranted to determine whether active standing could be a useful model to reduce the risk of poor health outcomes (e.g., obesity, diabetes, mortality, and cardiovascular disease). The analytical approach presented could specifically be of use in randomized control trials to test whether encouraging/discouraging specific types of standing leads to measurable improvements in physiological outcomes.
Consideration of Public Health Recommendations on Standing
Standing is increasingly recognized as an important health behavior and has recently been incorporated into public health guidelines, though not globally. While the 2020 Canadian 24-hr Movement Guidelines recommend “several hours of light-intensity physical activities, including standing” (Ross et al., 2020), standing behaviors have yet to be recognized as an independent behavior with unique health consequences, despite a large portion of waking hours being attributable to this behavior (Hamer & Stamatakis, 2020). The development of movement guidelines is largely based on population-wide observational studies that implement self-reported or device-based measures of habitual activity time (DiPietro et al., 2020). As such, the ability to distinguish between passive and active standing behaviors offers a unique opportunity to better understand the health impacts and, therefore, contribute to the conflicting evidence based on whether standing time is beneficial (Katzmarzyk et al., 2009) or not (Waters & Dick, 2015). While standing is incorporated into the Canadian Movement Guidelines, it is based on low-quality evidence (Katzmarzyk et al., 2009) but is included to support public health initiatives that advocate for engaging in quiet standing in place of sitting (Ross et al., 2020). Applying the thigh-worn accelerometry and the machine learning model from Kowalsky et al. (2024) in free-living environments for studies examining the relation between standing time and health outcomes could yield higher quality evidence. This, in turn, could provide better evidence as to whether guidelines should focus on specific standing principles (e.g., frequency, intensity, type, and duration) in international activity recommendations.
Conclusions
The study by Kowalsky et al. (2024) contributes to the measurement and classification of standing behaviors and their health implications. By developing a machine learning model that differentiates between active and passive standing using accelerometry data, the research provides a foundation for more precise assessment of free-living postural patterns. This can help inform public health guidelines and interventions aimed at understanding whether standing, or specific standing types, is suitable behavior to be promoted to yield health benefits. Future research should focus on refining these models, exploring the use of multiple monitor configurations, and validating these findings in free-living environments. These advancements have the potential to enhance our understanding of standing behaviors’ health impacts and the possible divergent physiological effects of active versus passive standing (Figure 1).
Visual summary of commentary.
Citation: Journal for the Measurement of Physical Behaviour 7, 1; 10.1123/jmpb.2024-0033
References
Andersen, J.H., Haahr, J.P., & Frost, P. (2007). Risk factors for more severe regional musculoskeletal symptoms: A two‐year prospective study of a general working population. Arthritis & Rheumatism, 56(4), 1355–1364.
Anderson, J., Granat, M.H., Williams, A.E., & Nester, C. (2019). Exploring occupational standing activities using accelerometer-based activity monitoring. Ergonomics, 62(8), 1055–1065.
Bach, K., Kongsvold, A., Bårdstu, H., Bardal, E.M., Kjærnli, H.S., Herland, S., Logacjov, A., & Mork, P.J. (2022). A machine learning classifier for detection of physical activity types and postures during free-living. Journal for the Measurement of Physical Behaviour, 5(1), 24–31.
Balasubramanian, V., Adalarasu, K., & Regulapati, R. (2009). Comparing dynamic and stationary standing postures in an assembly task. International Journal of Industrial Ergonomics, 39(5), 649–654.
Basset, D.R., John, D., Conger, S.A., Rider, B.C., Passmore, R.M., & Clark, J.M. (2014). Detection of lying down, sitting, standing, and stepping using two Activpal Monitors. Medicine & Science in Sports & Exercise, 46(10), 2025–2029.
Betts, J.A., Smith, H.A., Johnson-Bonson, D.A., Ellis, T.I., Dagnall, J., Hengist, A., Carroll, H., Thompson, D., Gonzalez, J.T., & Afman, G.H. (2019). The energy cost of sitting versus standing naturally in man. Medicine & Science in Sports & Exercise, 51(4), 726–733.
DiPietro, L., Al-Ansari, S.S., Biddle, S.J.H., Borodulin, K., Bull, F.C., Buman, M.P., Cardon, G., Carty, C., Chaput, J.-P., Chastin, S., Chou, R., Dempsey, P.C., Ekelund, U., Firth, J., Friedenreich, C.M., Garcia, L., Gichu, M., Jago, R., Katzmarzyk, P.T., . . . Willumsen, J.F. (2020). Advancing the global physical activity agenda: Recommendations for future research by the 2020 WHO physical activity and sedentary behavior guidelines development group. International Journal of Behavioral Nutrition and Physical Activity, 17(1), Article 143.
Drury, C.G., Hsiao, Y.L., Joseph, C., Joshi, S., Lapp, J., & Pennathur, P.R. (2008). Posture and performance: Sitting vs. standing for security screening. Ergonomics, 51(3), 290–307.
Hamer, M., & Stamatakis, E. (2020). The descriptive epidemiology of standing activity during free-living in 5412 middle-aged adults: The 1970 British Cohort Study. Journal of Epidemiology and Community Health, 74(9), 757–760.
Hwang, C.-L., Chen, S.-H., Chou, C.-H., Grigoriadis, G., Liao, T.-C., Fancher, I.S., Arena, R., & Phillips, S.A. (2022). The physiological benefits of sitting less and moving more: Opportunities for future research. Progress in Cardiovascular Diseases, 73, 61–66.
Jingjie, W., Yang, L., Jing, Y., Ran, L., Yiqing, X., & Zhou, N. (2022). Sedentary time and its association with risk of cardiovascular diseases in adults: An updated systematic review and meta-analysis of observational studies. BMC Public Health, 22(1), Article 286.
Katzmarzyk, P.T., Church, T.S., Craig, C.L., & Bouchard, C. (2009). Sitting time and mortality from all causes, cardiovascular disease, and cancer. Medicine & Science in Sports & Exercise, 41(5), 998–1005.
Kowalsky, R.J., van Werkhoven, H., Meucci, M., Quinn, T.D., Stoner, L., Hearon, C.M., & Barone Gibbs, B. (2024). Distinguishing passive and active standing behaviors from accelerometry. Journal for the Measurement of Physical Behaviour, 7(1), Article jmpb.2024-0004.
Levine, J.A., Schleusner, S.J., & Jensen, M.D. (2000). Energy expenditure of nonexercise activity. The American Journal of Clinical Nutrition, 72(6), 1451–1454.
Martin, C.J.H., Kenney, L., Pratt, T., & Granat, M.H. (2015). The development and validation of an activity monitoring system for use in measurement of posture of childbearing women during first stage of labor. Journal of Midwifery & Women’s Health, 60(2), 182–186.
Patterson, R., McNamara, E., Tainio, M., de Sá, T.H., Smith, A.D., Sharp, S.J., Edwards, P., Woodcock, J., Brage, S., & Wijndaele, K. (2018). Sedentary behaviour and risk of all-cause, cardiovascular and cancer mortality, and incident type 2 diabetes: A systematic review and dose response meta-analysis. European Journal of Epidemiology, 33(9), 811–829.
Ross, R., Chaput, J.-P., Giangregorio, L.M., Janssen, I., Saunders, T.J., Kho, M.E., Poitras, V.J., Tomasone, J.R., El-Kotob, R., McLaughlin, E.C., Duggan, M., Carrier, J., Carson, V., Chastin, S.F., Latimer-Cheung, A.E., Chulak-Bozzer, T., Faulkner, G., Flood, S.M., Gazendam, M.K., . . . Tremblay, M.S. (2020). Canadian 24-Hour Movement Guidelines for Adults aged 18–64 years and Adults aged 65 years or older: An integration of physical activity, sedentary behaviour, and sleep. Applied Physiology, Nutrition, and Metabolism, 45(10, Suppl. 2), S57–S102.
Stoner, L., Barone Gibbs, B., Meyer, M.L., Fryer, S., Credeur, D., Paterson, C., Stone, K., Hanson, E.D., Kowalsky, R.J., Horiuchi, M., Mack, C.P., & Dave, G. (2021). A primer on repeated sitting exposure and the cardiovascular system: Considerations for study design, analysis, interpretation, and translation. Frontiers in Cardiovascular Medicine, 8, Article 716938.
Taraldsen, K., Askim, T., Sletvold, O., Einarsen, E.K., Grüner Bjåstad, K., Indredavik, B., & Helbostad, J.L. (2011). Evaluation of a body-worn sensor system to measure physical activity in older people with impaired function. Physical Therapy, 91(2), 277–285.
van der Berg, J.D., Stehouwer, C.D.A., Bosma, H., van der Velde, J.H.P.M., Willems, P.J.B., Savelberg, H.H.C.M., Schram, M.T., Sep, S.J.S., van der Kallen, C.J.H., Henry, R.M.A., Dagnelie, P.C., Schaper, N.C., & Koster, A. (2016). Associations of total amount and patterns of sedentary behaviour with type 2 diabetes and the metabolic syndrome: The Maastricht Study. Diabetologia, 59(4), 709–718.
Waters, T.R., & Dick, R.B. (2015). Evidence of health risks associated with prolonged standing at work and intervention effectiveness. Rehabilitation Nursing, 40(3), 148–165.
Wu, Y., O’Brien, M.W., Peddle, A., Daley, W.S., Schwartz, B.D., Kimmerly, D.S., & Frayne, R.J. (2024). Criterion validity of accelerometers in determining knee-flexion angles during sitting in a laboratory setting. Journal for the Measurement of Physical Behaviour, 7(1), Article jmpb.2023-0027.
