Wearable with different characteristics making the choice of the

Wearablemotion sensors (accelerometers and pedometers) are popular tools for objectiveassessment. Pedometers are used to measure steps and distance whileaccelerometers measure acceleration and movement (Strathet al., 2013). Pedometers are motion sensors that record movement interms of steps taken. Early forms of pedometers used mechanical sensors thatidentified steps based on the force generated during walking. Nowadays, withthe advancement of technologies, they use microelectromechanical systems toidentify steps which considerably increased their accuracy. Most of them arehip-worn, but it is suggested that the more accurate position should be theankle.

Furthermore, some recent models also allow measurement of energyexpenditure, acceleration, and sleep (Plasqui, Bonomi& Westerterp, 2013). Accelerometers provide information about type,frequency, intensity, and duration of physical activity, and, thus, they arecommonly used in research studies. Similar topedometers, they are typically hip-worn, but can also be fixed to ankles orwrists. It is proposed that the more accurate position to wear accelerometersis the lower back or hip, i.e.

closer to the centre of the mass. They rely onmicroelectromechanical systems to record acceleration and objectively capturebody movements. Thanks to the technology advancements, they can detect types ofphysical activity and energy expenditure. There are many commercially availableaccelerometers with different characteristics making the choice of the mostsuitable accelerometer very difficult.

(Plasqui et al.,2013; Ainsworth et al., 2015)Notonly that the use of accelerometers increases in recent years, but withtechnology improvements, there is a tendency to insert them into smartphones as they are regularly used in everyday lives,especially among adolescents. It is proposed that designed application formobile phones should be used with other objective assessment monitors, whichwill improve the quality of collected data (Dunton etal., 2014; Shoaib, Bosch, Incel, Scholten & Havinga, 2014).

On the otherhand, there is a high inaccuracy of smartphone pedometer applications, whichsuggests caution in the interpretation of smartphone application data (Orr etal., 2015). Advantagesand disadvantages of motion sensorsTheuse of wearable motion sensors, such as pedometers and accelerometers, inphysical activity assessment increases in research and clinical assessment.However, the choice of the most adequate monitor will depend on severalfactors: research goal, target population, physical activity characteristics,cost-efficiency, and required measurement precision (Ainsworthet al., 2015).

Pedometersare inexpensive and present a low burden for participants. Further, they can beused in studies with many participants and data obtained from pedometers areeasily processed.  But pedometers do notmeasure intensity or duration of physical activity and are not accurate forassessment of energy expenditure (Strath et al., 2013).Pedometers alsofail to be accurate at slower walking speeds or when worn at pockets or wristsand they cannot detect sedentary activities, posture, and energy expenditure(Ainsworth et al., 2015).Advantagesof using accelerometers include detailed data about intensity, frequency, andduration of physical activity, they are relatively inexpensive, small, andnon-invasive.

The memory capacity increases nowadays, so data can be collectedover longer period of time. However, they are not suitable for all physicalactivities, especially those that require the activity of the upper body parts.Also, data are not measured in commonly used units and transformation of unitsis time demanding (Strath et al., 2013). One ofthe important advantages of accelerometers is the possibility to detect seatedpostures and transitions between seated and standing postures.

Yet, only few ofthem can measure light-intensity physical activity and sedentary behaviour(Ainsworth et al., 2015).Thereis a number of motion sensors commercially available for the assessment ofphysical activity. Plasqui et al. (2013) compared validity of accelerometersused in 15 different validation studies and proposed the necessity ofvalidation of accelerometers against doubly labelled water method. Althoughaccelerometers provide daily data in the assessment of physical activity anddoubly labelled water provides a measure of energy expenditure over a period oftime and both methods are prone to the error, for the most accurate measures ofphysical activity both methods should be used complementarily (Plasqui et al.

, 2013).In the study of Lee et al. (2014), eight different typesof motion sensors were investigated for the accuracy to estimate energyexpenditure. Participants wore all of them at the same time during activityroutine of 13 different activities categorized into sedentary, walking, runningand moderate-to vigorous activities.

Devices were validated against ActiGraph, as the one most commonly used and almostall of them showed good potential for the assessment of physical activity (Lee, Kim and Welk, 2014).Thetechnology development provides opportunities to improve physical activityassessment methods and overcome disadvantages of current methods. Pedometersand most of accelerometers detect movements in the vertical plane.

But someaccelerometers are sensitive to two or three planes and able to detectdifferent physical activities (McCarthy & Grey,2015). Triaxial accelerometers show a highsensitivity for sitting, standing, walking, running, and cycling (Skotte,Korshøj, Kristiansen, Hanisch & Holtermann, 2014). Gatti et al. (2015)found excellent reliability and validity of a triaxial accelerometer placed atthe waist and shank during running and pedal-revolution counts during bicycling(Gatti, Stratford, Brenneman & Maly, 2015). They also have apotential to be used to measure upper extremity physical activity, especiallyif worn on wrists.

That way they monitor arm usage and even detect differencesin slow arm movements, suggesting the importance in their usage duringrehabilitation (Lawinger, Uhl, Abel & Kamineni,2015). Still, Pediši? and Bauman (2014) suggest that the use of motion sensorsis general population studies is still limited due to different study designs,validity, between-study comparability and simplicity. Furtherproblem that could occur with motion sensors is limitation in cooperation withparticipants. Participants could easily forget or refuse to wear them, and theyusually remove them during sleep and water-related activities (Dunton et al.

,2014). Theuse of motion sensors in clinical studiesSedentarybehaviour increases the risk of chronic diseases and it is now identified asone of the leading causes of global mortality. For this reason, physicalactivity has important benefits in the general population and the World HealthOrganisation (WHO) recognises its importance in health. Research in this areaprovides important information about the dose-response relationship betweenphysical activity and health. This, together with the valid methods for theassessment of physical activity, offers necessary information to make anintervention plan to reduce sedentary behaviour (WHO, 2010). It is required toaddress physical inactivity and develop specific interventions and implementthem at the national levels to increase physical activity among population and,thus, decrease the burden of disease (Bauman, Merom,Bull, Buchner & Fiatarone Singh, 2016).

Understandingthe consequence of lifestyle and not only genetic factors in the development ofmany diseases, current recommendations for their prevention include physicalactivity. Motion sensors can be used to examine at which levels physicalactivity can affect metabolic changes indiabetic patients and be clinically beneficial (Herziget al. 2013). By using a motion sensor among patients with diabetes, lowlevels of physical activity in patients, in term of total energy expenditure,number of steps, and duration of physical activity, are observed (Fagour et al., 2013). Similarly, low levels physicalactivity are detected among people with depressive and anxiety disorders,measured by using accelerometer. Grounding the results on accelerometermeasures, it is recommended that for this type of patients, light physicalactivity is more efficient than high-intensity physical activity in reducingthe disorders manifestation (Helgadóttir, Forsell &Ekblom, 2015).

By recognizing the consequences of sedentary behaviourin the development of diseases and the importance of physical activity toimprove health outcomes, motion sensors become very important monitoring andinterventional tools. It is reported that they can be used as intervention toimprove glucose metabolism with increase in physical activity in diabeticpatients (Miyazaki & Kotani, 2015).Pedometer-driven physical activity is used as an intervention to increasephysical activity and consequently improve health. This is confirmed forseveral diseases, such as diabetes (Guglani, Shenoy andSandhu, 2014), obesity (Cai et al., 2016),mental illness (Helgadóttir et al.

, 2015), musculoskeletaldiseases (Mansi et al., 2014), and chronicobstructive pulmonary disease (Mendoza et al., 2014).

Still, future studies are required for further clarification. ConclusionsByunderstanding the effect of physical inactivity on health, there is a need forvalidated methods that measure physical activity and inactivity. There is nogold standard for motion sensors and the choice of the optimal motion sensor iscomplex. Motion sensors eliminate the problems of subjective methods, but theyare more money and time consuming and as discussed, they have their own(dis)advantages.

Motion sensors have the advantage of cost, non-invasivenessand clear data. Still, there are lot limitations and it is suggested to usethem simultaneously with other physical assessment methods to improve the dataquality. A large heterogeneity in assessment of different types of motionsensors across studies exists and data need to be interpreted with a caution. Yet,they provide very important data in clinical studies. Not only that motionsensors can be used in monitoring, but also in health intervention.

The validinterpretation of data in these studies can help in minimizing sedentarybehaviour and improve the assessment of health outcomes associated withincreased physical activity. Further research is necessary to support the useof motion sensors interventions as long term interventions for chronicdiseases.