A school-based intervention improves physical fitness in Ecuadorian adolescents: a cluster-randomized controlled trial

Table 5 Subgroup analysis of physical fitness, screen time, physical activity according to school characteristics

Outcome	n	Control group	Intervention group	Separate model per school characteristic		Full model adjusted for school characteristics
		Mean (SD)	Mean (SD)	Difference [95% CI] ^a	P	Difference [95% CI] ^b	P
Speed shuttle run 10×5 (s)
School size	1021				<0.01^c		<0.01^c
≤695 students	476	1.58 (3.59)	1.93 (2.47)	0.30 [−0.55-1.14]	0.25	0.22 [−0.36 -0.80]	0.24
>695 students	545	3.63 (3.02)	1.85 (2.32)	−1.83 [−2.34 to −1.32]	<0.01	−1.49 [−2.20 to −0.78]	<0.01
Vertical jump (cm)
School gender	1038				0.03^c		0.03^c
Male and female students	698	0.29 (6.82)	2.93 (6.70)	3.55 [1.73-5.36]	<0.01	3.57 [1.76-5.38]	<0.01
Only female students	340	−0.34 (5.67)	−0.30 (6.48)	0.14 [−2.21-2.49]	0.45	0.06 [−2.16-2.28]	0.48
Flamingo (trying/min)
Physical activity space	571				0.04^c		0.04^c
≤4.07 (students/m²)	237	−5.26 (7.87)	−0.82 (6.40)	3.29 [0.97 - 5.61]	0.01	3.29 [0.97 - 5.61]	0.01
>4.07 (students/m²)	334	−3.38 (7.38)	−2.43 (6.68)	0.91 [−0.99 - 2.81]	0.19	0.91 [−0.99 - 2.81]	0.19
% sedentary in weekends (screen time >3 h/day)
School type	1071				0.03^c		0.05^c
Private	438	0.18 (0.39)	0.31 (0.46)	0.13 [0.04 - 0.21]	<0.01	0.13 [0.03 - 0.23]	0.01
Public	633	0.26 (0.44)	0.25 (0.43)	−0.02 [−0.10 - 0.05]	0.29	0.00 [−0.14 - 0.13]	0.48
Proportion of adolescents who meet the recommendation (60 min of MVPA/day)
School schedule	134				0.01^c		0.02^c
Half-day class	110	0.70 (0.46)	0.96 (0.19)	0.28 [0.14-0.42]	<0.01	0.37 [0.14 - 0.61]	0.01
Full-day class	24	1.00 (0.00)	0.77 (0.44)	−0.29 [−0.54 to −0.04]	0.02	−0.29 [−0.54 to −0.04]	0.02

^aDifferences and CI were obtained from a linear mixed model adjusted for BMI z-scores, gender, socio-economic status and physical activity knowledge at baseline, including the interaction term between school characteristic and allocation group.
^bDifferences and CI were obtained from linear mixed models adjusted for BMI z-scores, gender, socio-economic status and physical activity knowledge, including the interaction term between school characteristic and allocation group. Models included all school characteristics as fixed effects that were significantly associated with the outcome (P < 0.05) and all significant interaction terms (P < 0.1) from the separate models.
^cP for interaction (significant at P < 0.10).
CI: confidence interval; SD: standard deviation.

ISSN: 1479-5868