An investigation into the optimal wear time criteria necessary to reliably estimate physical activity and sedentary behaviour from ActiGraph wGT3X+ accelerometer data in older care home residents

Airlie, Jennifer; Forster, Anne; Birch, Karen M.

doi:10.1186/s12877-021-02725-6

Table 5 Accelerometer outcomes calculated based on an increasing number of days of data collection and equivalence results

From: An investigation into the optimal wear time criteria necessary to reliably estimate physical activity and sedentary behaviour from ActiGraph wGT3X+ accelerometer data in older care home residents

Counts.day⁻¹
Day (n)	1	2	3	4	5	6	7 (EI: ± 14,151)
Mean ± SD	124,046 ± 85,406	140,581 ± 94,191	137,584 ± 97,130	136,500 ± 96,012	136,262 ± 93,744	139,427 ± 98,997	141,510 ± 104,549
Mean difference^b	17463^a	929	3926	5010	5248	2082	n/a
Effect Size	0.18	0.01	0.04	0.05	0.05	0.02	n/a
90% CI around mean difference
Lower	3295	−13,239	−10,242	− 9158	− 8920	− 12,085	n/a
Upper	31,631	15,097	18,093	19,177	19,415	16,250	n/a
Counts^.minute⁻¹ (cpm)
Day (n)	1	2	3	4	5	6	7 (EI: ± 18)
Mean ± SD	181 ± 123	186 ± 129	174 ± 123	171 ± 124	170 ± 119	172 ± 123	172 ± 125
Mean difference^b	9	14	2	1	2	0	n/a
Effect Size	0.08	0.11	0.02	0.01	0.02	0.00	n/a
90% CI around mean difference
Lower	- 24	- 28	−17	- 13	- 12	- 14	n/a
Upper	4^b	1^b	12^b	15^b	17^b	14^b	n/a
PA time (h and min)
Day (n)	1	2	3	4	5	6	7 (EI: ± 18 min)
Mean ± SD	2 h 27 min ± 1 h 34 min	2 h 44 min ± 1 h 40 min	2 h 45 min ± 1 h 46 min	2 h 45 min ± 1 h 46 min	2 h 45 min ± 1 h 44 min	2 h 48 min ± 1 h 45 min	2 h 50 min ± 1 h 46 min
Mean difference^†	23 min^a	6 min	5 min	5 min	5 min	2 min	n/a
Effect Size	0.23	0.06	0.05	0.04	0.04	0.02	n/a
90% CI around mean difference
Lower	9 min	- 8 min	- 9 min	- 8 min	- 9 min	- 12 min	n/a
Upper	37 min	20 min	19 min	19 min	19 min	16 min^b	n/a
SB time (h and min)
Day (n)	1	2	3	4	5	6	7 (EI: ± 1 h 4 min)
Mean ± SD	9 h 3 min ± 2 h 2 min	9 h 50 min ± 2 h 2 min	10 h 15 min ± 1 h 51 min	10 h 28 min ± 1 h 51 min	10 h 34 min ± 1 h 51 min	10 h 35 min ± 1 h 49 min	10 h 38 min ± 1 h 45 min
Mean difference^b	1 h 35 min^a	48 min^a	23 min	10 min	5 min	3 min	n/a
Effect Size	0.83	0.44	0.21	0.09	0.04	0.03	n/a
90% CI around mean difference
Lower	1 h 6 min	19 min	- 6 min	- 19 min	- 25 min	- 27 min	n/a
Upper	2 h 4 min	1 h 17 min	52 min^b	39 min^b	34 min^b	32 min^b	n/a

Notes. Accelerometer outcomes are: counts.day^− 1, counts.minute^− 1, PA time and SB time. A valid day being defined as a having a wear time of ≥ 8 h. Mean difference from the seven-day reference point. ^a denotes that the mean difference from the estimate based on 7 days is significant at the 0.10 level. Bonferroni correction applied. Equivalence interval = ± 10% of mean when the minimum daily WT of ≥10 h was used. For counts.day^− 1 = 14,151 count.day^− 1; cpm = 18 cpm; PA time = 18 min; SB time = 1 h 4 min. ^b indicates equivalency
Abbreviations: CI confidence interval; PA physical activity; SB sedentary behaviour; cpm counts per minute

Back to article page

ISSN: 1471-2318

Contact us

Submission enquiries: bmcgeriatrics@biomedcentral.com
General enquiries: ORSupport@springernature.com

BMC Geriatrics

Contact us