New relative intensity ambulatory accelerometer thresholds for elderly men and women: the Generation 100 study

Public health initiatives world-wide recommend increasing physical activity (PA) to improve health. However, the dose and the intensity of PA producing the most benefit are still debated. Accurate assessment of PA is necessary in order to 1) investigate the dose–response relationship between PA and health, 2) shape the most beneficial public health initiatives and 3) test the effectiveness of such initiatives. Actigraph accelerometer is widely used to objectively assess PA, and the raw data is given in counts per unit time. Count-thresholds for low, moderate and vigorous PA are mostly based on absolute intensity. This leads to largely inadequate PA intensity assessment in a large proportion of the elderly, who due to their declining maximal oxygen uptake (VO_2max) cannot reach the moderate/vigorous intensity as defined in absolute terms. To resolve this issue, here we report relative Actigraph intensity-thresholds for the elderly.

Submaximal-oxygen-uptake, VO_2max and maximal heart rate (HR_max) were measured in 111 70–77 year olds, while wearing an Actigraph-GT3X+. Relationship between VO_2max percentage (%), counts-per-minute (CPM) and gender (for both the vertical-axis (VA) and vector-magnitude (VM)) and VO_2max% and HR_max% was established using a mixed-regression-model. VM-and VA-models were tested against each other to see which model predicts intensity of PA better.

VO_2max and gender significantly affected number of CPM at different PA intensities (p < 0.05). Therefore, intensity-thresholds were created for both men and women of ranging VO_2max values (low, medium, high). VM-model was found to be a better predictor of PA-intensity than VA-model (p < 0.05). Established thresholds for moderate intensity (46−63 % of VO_2max) ranged from 669–3367 and 834–4048 CPM and vigorous intensity (64−90 % of VO_2max) from 1625–4868 and 2012-5423CPM, for women and men, respectively. Lastly, we used this evidence to derive a formula that predicts customized relative intensity of PA (either VO_2max% or HR_max%) using counts-per-minute values as input.

Intensity-thresholds depend on VO_2max, gender and Actigraph-axis. PA intensity-thresholds that take all these factors into account allow for more accurate relative intensity PA assessment in the elderly and will be useful in future PA research.

(ClinicalTrials.gov Identifier: NCT02017847, registered 17. December 2013)