Table 3.

Results of overall, sensitivity and subgroup analyses

Type of analysis	N. of included datasets	ES	95% CI, p-value	N. of participants	χ2; df	I2	p-value	Intercepta	t-valuea	p-valuea
Overall	60	0.83	(0.74; 0.93), 0.001	557 111	895.19; 59	93.41	<0.001	0.53	0.78	0.439
Sensitivity analysis
QS ⩾15	47	0.79	(0.68; 0.91), 0.001	485 092	808.42; 46	94.31	<0.0001	0.52	0.65	0.520
QS ⩾15 + validated tool to diagnose depression	44	0.76	(0.65; 0.88), 0.0001	239 453	763.78; 43	94.37	<0.0001	0.33	0.41	0.687
QS ⩾15+ validated tool to diagnose depression + longitudinal design	24	0.76	(0.64; 0.90), 0.001	162 004	652.18; 23	96.47	<0.001	0.85	0.52	0.607
Subgroup analysis by study design
Longitudinal	26	0.79	(0.67; 0.93), 0.004	407 086	681.14; 25	96.33	<0.001	1.18	0.76	0.455
Cross-sectional	33	0.89	(0.76; 1.04), 0.136	139 484	161.43; 32	80.18	<0.001	−0.24	−0.48	0.638
Subgroup analysis by gender
Women	21	0.79	(0.61; 1.02), 0.074	219 655	189.48; 20	89.44	<0.001	−0.25	−0.25	0.805
Men	20	0.87	(0.68; 1.11), 0.258	223 840	252.80; 19	92.48	<0.001	0.99	0.93	0.366

df, degree of freedom; ES, Effect Size; N., number; QS, quality score

^aEgger's linear regression test.