GGIR thresholds for hip-worn ActiGraph gt3x data in toddlers aged 18 to 30 months

[LoreDaelman]

Hi!

I collected hip-worn ActiGraph gt3x data in toddlers aged 18 to 30 months. They wore it for 7 consecutive days. They also wore the accelerometer during sleep at night and during naps, however I am currently using the sleeplogs since I don't think there is a sleep algorithm for these young children.

I noticed that there are no valid GGIR threshholds for toddlers yet. Can I still use GGIR? Or should I use the ActiLife software with epochs and the Trost et al. cutpoints?

Thank you in advance!

Kind regards,

Lore

[vincentvanhees]

GGIR has many parameters that could be considered thresholds. I assume you are referring to acceleration thresholds (cut-points) to distinguishing level of intensity.

• The overview lists cut-points for the wrist: https://wadpac.github.io/GGIR/articles/CutPoints.html#cut-points-for-preschoolers. Note that I am not actively maintaining the overview, so you may want to check the literature as well.
• It is possible with GGIR to convert data to ActiGraph-'counts' and then use thresholds proposed for those via GGIR.

More generally speaking, I do not think it makes much sense to aim for distinct intensity levels in this young age group: It is difficult or even impossible to get infants to do any behaviour consistently in a lab environment in way that steady state energy expenditure is reached, which we need to derive reliable thresholds. I would be highly suspicious of studies that propose cut-points for this age groups.

Instead I would advise exploring alternative ways of describing behaviour, and for that GGIR offers plenty of outcome variables.