Table 2.

Effect of age at commencement of involvement in talent promotion programs on short-term junior performance and long-term senior performance^a

Subsample	$\overline{d}$	95% CI	k	p value	I2
Junior athletes
Overall	− 0.53	− 0.74 to − 0.32	16	< 0.001	74.94
Talent promotion programs
Age at entry into federation’s squadb	− 0.64	− 0.76 to − 0.52	8	< 0.001	3.55
Age at entry into youth sport academyb	− 0.42	− 0.83 to − 0.01	8	0.044	82.15
Performance levels
World-class versus lowerc	− 0.63	− 0.76 to − 0.49	7	< 0.001	13.97
National or regional class vs lowerc	− 0.46	− 0.84 to − 0.09	9	0.016	79.57
Types of sports
Individual sports	− 0.59	− 0.78 to − 0.40	6	< 0.001	36.11
Team sports	− 0.46	− 0.84 to − 0.09	9	0.016	79.57
Senior athletes
Overall	0.56	0.40–0.72	34	< 0.001	74.83
Talent promotion programs
Age at entry into federation’s squadb	0.57	0.39–0.74	29	< 0.001	75.50
Age at entry into youth sport academyb	0.53	0.13–0.93	5	0.009	71.15
Performance levels
World-class vs lowerc	0.65	0.44–0.85	23	< 0.001	73.98
National or regional class vs lowerc	0.42	0.19–0.65	11	< 0.001	66.58
Types of sports
Individual sports	0.50	0.33–0.66	11	< 0.001	12.02
Team sports	0.57	0.35–0.79	20	< 0.001	81.11

CI confidence interval. The forest plots for junior and senior samples are shown in the Figs. S1 and S2 of the ESM

^aNote the sign of effects: a positive effect indicates that higher performance was associated with older age at commencement of TPP involvement. A negative effect indicates that higher performance was associated with younger age at commencement of TPP involvement

^bInvolving all performance levels

^cInvolving ages at entry into federation’s squads and youth sport academies