The importance of accelerometry and gyroscope for eating behaviour and associated intake

The article "OREBA: A Dataset for Objectively Recognizing Eating Behaviour and Associated Intake“ (Rouast et al., 2020) shows a comprehensive multi-sensor recording of communal intake occasions for researchers interested in automatic detection of intake gestures.
Modern multi-sensors like the Move 4 provide researchers the ability to collect objectively a large dataset regarding the accelerometry and gyroscope-data. That is very important for automatic detection of intake gestures that is a key element of autonomic dietary monitoring. Read more and klick the article above.

Mobile Sensing

movisensXS provides researchers easy access to the Mobile Sensing features of smartphones. The additional insights derived through Mobile Sensing add context to questionnaire data, and can help determine optimal times for interventions. This is useful for researchers of Interactive Ambulatory Assessment. Learn more about the possibilities of Mobile Sensing and about Social Sensing and their application within Interactive Ambulatory Assessment in our recommendations and in our webinar.

Ecological Momentary Interventions with movisens

The solutions from movisens make Ecological Momentary Interventions realizable with movisensXS.

What triggers are possible
  • Questionnaire replies
  • Condition Mutable Values
  • Mobile Sensing
  • Sensor trigger
  • Complex analysis via analysis server (questionnaire answers, sensor data, also based on history)
  • Trigger from ExternerApp
How can interventions be designed
  • Text
  • Audio
  • Video
  • Generated feedback from the server
  • Gamification
  • Call external app

Learn more about the possibilities that movisens currently offers on EMIs

Study Diary VI

SedentaryMood-Study (Part VI)

The following article is part of a series about the "SedentaryMood-Study".

The practical implementation of the SedentaryMood-Study

In the last article the applied investigation plan was explained. The following article describes the practical implementation of the study.

Step by step

  • Preparation of the ethics proposal
  • Preparation of the respondent information and questionnaires
  • Creation of the study concept via the Ambulatory Assessment Platform movisensXS

  • Preparation of the necessary research equipment and the associated materials
  • Configuring and Starting Sensors

  • Install TriggerApp
  • Bluetooth low energy Establish connection between sensor and smartphone and select algorithm (Sedentary)

  • Pair the smartphone with movisensXS and load the created design on the smartphone
  • Instruction and instruction of test persons at the workplace
  • Start study!

  • you can find out more about the study in the next article...

    Study Diary V

    SedentaryMood-Study (Part V)

    The following article is part of a series about the "SedentaryMood-Study".

    Measurement times of the real-time study

    In the last article, the sedentary sedimentary triggered e diaries and the randomly selected queries were described. In the next step, the study plan used for the SedentaryMood-Study is explained. Here, the frequency of the mood polls during the course of a day plays an important role.

    The exact number of queries depends on the individual participant's level of activity and thus, as in this study, up to 12 queries per day can be expected. More mood queries per day over a longer period of time are not to be recommended in order not to overstrain the associated willingness of the test persons to participate and thus not to endanger the data quality.

    On the basis of study results, about three to five days - of which at least one weekend day - are necessary for a representative recording of sedentary behaviour. In view of this, the survey period, the SedentaryMood Study, lasted five days (three working days and two weekend days).

    The sample was recruited from the University of Newcastle (UoN, Australia) and the Karlsruhe Institute of Technology (KIT, Germany). Participation in the study was linked to the following inclusion criteria: official co-worker of the institution, no illness or injury, and the work was performed predominantly in a sedentary body position.

    more about the practical implementation of the study can be found in the next article...

    Study diary IIII

    SedentaryMood Study (Part IIII)

    The following article is part of a series about the "SedentaryMood-Study".

    Sedentary Triggered e-Diaries

    The methodical core of the "SedentaryMood-Study" is the use of Sedentary Triggered e-Diaries. As far as is known, no comparable technical solution has been implemented in any study to date.

    Smartphone mit movisensXS
    Smartphone with movisensXS
    In this case, the application aims to capture psychological parameters precisely during long sitting phases. Sedentary behaviour is recorded, analysed in real time and mood queries are generated when defined thresholds are reached. The threshold for long sedentary behaviour is 30 minutes. The communication between smartphone (e-Diary) and accelerometer takes place via a Bluetooth low energy (BLE) connection.
    In the practical implementation of this study, Sedentary Triggerd e-Diaries as well as random time points in the daily routine were selected. In addition, at the end of a study day, the test persons were encouraged to subjectively indicate the entire duration of the completed day.

    The figure shows a triggered query with movisensXS.

    You can find out more about the SedentaryMood-Study in the next article...

    Sedentary Mood Study – Part 3

    Sedentary Mood Study (Part III)

    The following article is part of a series on the "SedentaryMood Study".

    Preparations to determine the body position

    As part of the "SedentaryMood Study," the Move 3 activity sensor was used for the first time to capture sedentary behavior as a primary target variable. Thus, some preparatory work was necessary in the run-up to the study. These were aimed primarily at the detection of the body position.

    In order to determine the most suitable carrying position for the activity sensor and thus to differentiate between sitting and upright body positions, two video studies were carried out. On the one hand in everyday life with different activities and body positions and on the other hand during office work in everyday working life.

    The evaluations of both measurements show that a carrying position of the activity sensor Move 3 between the sagittal axis and the longitudinal axis, an upright body position is assumed. If the sensor is in a position that exceeds the angular range of < 20° between the sagittal axis and the longitudinal axis, an upright body position is assumed. If the sensor is in a position that exceeds the angular range of > to the sagittal axis, then a seated / lying body position is specified.

    As the measured data of the activity sensor Move 3 analyzed, coupled with the e-Diary movisensXS and triggered with the SensorTrigger you will learn in the next article ...

    Sedentary Mood – Study II

    Study Diary "Sedentary Mood-Study"

    Sedentary Behavior Study

    The following article continues our series on "Sedentary Mood-Studies". Throughout the series we'll take you through a course of sedentary behavior studies, describing the process from planning to results. We'll start with the process of capturing the necessary data to examine the link between sedentary behavior and mood. To do this, we'll detail a particular sedentary behavior study to illustrate the idea.

    Part 2: Recording mood and sedentary behavior data

    When studying sedentary behavior, it's important to select a time frame that provides enough data to analyse. In this case, the researcher chose a five day period in an ambulatory setting to capture data in everyday life. Over this five day period participants received mood assessments several times per day, and had their activity recorded during waking hours.

    To capture the participants mood, it's necessary to use an experience sampling method. For this study, the android based experience sampling app movisensXS displayed the short version of the "Multidimensional Mood Questionnaire (MDMQ)" at random intervals. The MDMQ measures the mood in three dimensions - Valence, Energetic Arousal, and Calmness - and was specifically conceptualized for ambulatory studies (see Wilhelm and Schoebi, 2007, p. 259ff.)

    Whilst there's no technical device that captures sedentary behavior (see Kang and Rowe, 2015, p.113), the activity sensor serves as the de facto research instrument of choice. Given the abundance of fitness trackers on the market, it's often tempting to purchase inexpensive devices in order to obtain more data points. However, for research grade data it's important to use research grade devices. Whilst that may mean fewer devices and fewer participants, the quality of the data more than compensates.

    The Sedentary Behavior Research Network (SBRN, 2017) defines Sedentary Behavior as: "Sedentary behavior is any waking behavior characterized by an energy expenditure ≤1.5 metabolic equivalents (METs), while in a sitting, reclining or lying posture". So in order to obtain an optimal recording, we need to evaluate both body position and energy expenditure.

    In the next part of this series, we'll describe how the guidelines for sedentary behavior were met within this study

    “Sedentary Mood Study”

    Study Diary "Sedentary Mood Study"

    The following article begins a "Sedentary Behavior Studies" series, in which we'll detail the study process from conception through to results. We'll focus on the link between sedentary behavior and mood. But first, we need to address an important point.

    Why research sedentary behavior?

    "Whoever sits longer dies younger" and "Sitting is the new smoking". These headlines come from recent behavioral and health related studies. Due to such conclusions sedentary behavior demands the attention of researchers, and the general public looks on with interest. As the understanding of sedentary behavior and it's consequences develops, further research appears illustrating its adverse effects on cardiovascular health, metabolic and muscular parameters, and it's dire risk for overall physical health (Owen et al., 2010, p.3).

    Despite these findings, there's little research on the effect of sedentary behavior on psychological health. Preliminary findings indicate a possible connection between the two. As Fuchs et al. suggest (2015. p. 7) "Sitting changes the activity of metabolism and therefore it doesn´t seem absurd, that this can also influence the psychological processes." If the link between sedentary behaviour and adverse psychological health exists, understanding it would prove valuable in the fight against psychosomatic diseases.

    "How do we measure sedentary behavior and mood?" ... to be continued in the next articles

    Measuring “Sedentary Behavior” – General recommendations and solutions by movisens

    “Sedentary behavior” is defined as sitting or lying with low energy expenditure (SBRN, 2012). In isolation, energy expenditure provides no reliable information about body posture (e.g. sitting vs. standing), an essential component of assessing sedentary behavior (Holtermann, et al., 2017). Conversley, assessing body posture alone cannot provide insights into energy expenditure, e.g. machine sewing while sitting, crane operators, lifting weights in the gym (Holtermann, et al., 2017). To accurately determine sedentary behaviour requires the assessment of both body posture and energy expenditure.

    Reliable methods to measure and assess "sedentary behavior"

    • Case 1: Known environment – e.g. no possibility to stand and all low physical activity ≙ sitting. In this case attaching a sensor at the hip (Move 3) provides only a rough estimation of sedentary behavior.
    • Case 2: Differentiating between sitting/lying & standing. By attaching a sensor (Move 3) to the thigh, the different angles of the axis allow differentiation between sitting/lying and standing. But it is not possible to differentiate between sitting and lying (Byrom, Stratton, McCarthy, & Muelhausen, 2016).
    • Case 3: Assessing changes in time distribution of sedentary behavior – requiring the precise distribution of sedentary behavior and physical activity intensity. This case requires the assessment of both body posture and energy expenditure. This is possible by attaching one sensor (Move 3) at the thigh (sitting/lying vs. standing) and one (Move 3) at the upper body/hip (standing/sitting vs. lying). Additionally, the sensor at the upper body/hip provides the data necessary to estimate energy expenditure (Holtermann, et al., 2017)
    • Case 4: Assessing changes in time distribution of sedentary behavior with a static load. In this case the use of an ecg-sensor provides additional data to assess energy expenditure. Initially the ecg-sensor requires calibration to estimate energy expenditure with additional load. Attaching one sensor (Move 3) at the thigh (sitting/lying vs. standing) and one physical activity and ecg-sensor (EcgMove 3) at the upper body (standing/sitting vs. lying) provides acceleration data from two positions, and the additional ecg-signal allows improved energy expenditure estimations during static work, due to the linear relationship between cardiorespiratory stress and energy expenditure (Holtermann, et al., 2017).
    • Case 5: If an intervention is necessary, or if the research requires additional subjective parameters, we offer the possibility to trigger a questionnaire with our experience sampling platform movisensXS via our SensorTrigger. After the application detects 30 minutes (customizable by the researcher) of sedentary behaviour (<1.5MET) from the sensor, the trigger displays a form on a smartphone app prompting the participant to answer a questionnaire. This offers the possibility to obtain detailed feedback and insights into the daily routine of the study participant.