Table 2.

Subgroup analyses based on participant and study characteristics.

Characteristic	Number of studies	Meta-analysis		Heterogeneity		Between-group test
		Hedges's g	95% CI	I 2	p	Chi2	p
Participant characteristics
Mean age
Children and adolescents	1	—	—	—	—	1.78	0.18
Adults (18–60 years old)	18	−0.63	−1.08 to −0.17a	97%	<0.01
Older adults (above 60 years old)	3	−0.27	−0.53 to −0.00a	24%	0.27
Gender
Male < female	18	−0.59	−1.05 to −0.13a	97%	<0.01	0.71	0.70
Male > female	3	−0.34	−0.71 to 0.03	78%	0.01
No report	1	−0.44	−0.78 to −0.09a	—	—
Associated disease
With disease	7	−0.22	−0.36 to −0.08a	26%	0.23	2.91	0.09
No mention disease	15	−0.71	−1.25 to −0.17a	97%	<0.01
Study characteristics
Year of publication
10 years ago	8	−0.21	−0.39 to −0.03a	35%	0.15	3.08	0.08
Recent 10 years	14	−0.74	−1.31 to −0.18a	98%	<0.01
Country of publication
Asia	7	−1.13	−2.22 to −0.04a	99%	<0.01	7.19	0.07
Australia	1	−0.87	−1.44 to −0.30a	—	—
Europe	11	−0.27	−0.40 to −0.14a	53%	0.02
North America	3	−0.16	−0.45 to 0.14	0%	0.62
Types of CBT approaches
in-CBT	2	−2.87	−8.18 to 2.44	—	—	7.88	0.05
g-CBT	10	−0.11	−0.24 to 0.02	9%	0.36
i-CBT	12	−0.81	−1.38 to −0.25a	98%	<0.01
t-CBT	2	−0.43	−0.91 to 0.05	—	—
Intervention duration
<1 M	3	−0.24	−0.42 to −0.06a	0%	0.55	2.13	0.34
1 –3 M	14	−0.69	−1.27 to −0.11a	98%	<0.01
>3 M	5	−0.31	−0.45 to −0.16a	23%	0.27
Follow-up duration
FU < 6 M	4	−0.69	−0.90 to −0.48a	18%	0.30	1.85	0.40
6 ≤ FU < 12 M	6	−0.83	−2.03 to 0.37	98%	<0.01
FU ≥ 12 M	12	−0.39	−0.78 to 0.00	96%	<0.01
Attrition rate
High	9	−0.98	−1.83 to −0.12a	98%	<0.01	2.83	0.09
Low and moderate	13	−0.24	−0.36 to −0.12a	48%	0.03

Note: Justification for each subgroup type: age, characterized into three groups: children and adolescents (0–17), adults (18–60), and older adults (above 60) based on previous cutoff [8]; gender: male < female and based on previous review suggested that females are three times more likely to have subclinical depression [8]; associated disease: to see if the occurring symptoms from other diseases would influence the intervention outcome. Intervention duration, characterized into three groups: <1 M, 1–3 M, >3 M to ensure there are enough studies to conduct subgroup analysis; follow-up duration, characterized into three groups: 1 < FU ≤ 6 M, 6 < FU < 12, and ≥12 M due to the lack of studies conducted follow-up in less than 3 months, in 3 months, and 6 month which might result in low statistical power; attrition rate, characterized into two groups: high (≥30%) and low and moderate (<30%), the merge of low and moderate attrition was due to the lack of studies that have <10% attrition which might result in low statistical power.

^aRepresents statistically significant.