The study investigated the association of the age at commencement of TPP involvement with later performance. The central finding is that athletes’ age at commencement of TPP involvement had opposite effects on short-term junior performance and long-term senior performance. Higher-performing junior athletes commenced TPP involvement at younger ages than lower-performing junior athletes. In contrast, higher-performing senior athletes commenced TPP involvement at older ages than lower-performing senior athletes. The findings were robust, in terms of both direction and scale of effects, across federations’ squads and youth sport academies, performance levels compared (international, national, regional), as well as different types of sports.

The results are consistent with recent meta-analytical findings [42, 43, 63]: (1) successful junior athletes and successful senior athletes are not one identical population but are largely two disparate populations [63]. Most successful junior athletes achieve lower competition levels when they are seniors, while most successful senior athletes had achieved lower competition levels when they were juniors. The overlap of successful juniors and successful seniors is the smaller the higher the performance level and the younger the junior age category. (2) Concerning athletes’ participation patterns, several predictors of early junior performance and of long-term senior performance are opposite [42, 43]. Higher-performing junior athletes, compared with lower-performing juniors, started playing their respective main sport at younger ages, accumulated greater amounts of organized coach-led practice in their main sport and less practice in other sports, and achieved performance-related developmental ‘milestones’ at younger ages (e.g., first state, national, or international championships). In contrast, senior world-class athletes, compared with lower-performing senior national-class counterparts, started playing their main sport at older ages, accumulated less main-sport practice and more other-sports practice, and reached performance ‘milestones’ at older ages. It is important to note that across primary studies, the performance-related effects of the different predictors—main-sport starting age, amount of main-sport practice, amount of other-sports practice, and age of ‘milestone’ achievement—were closely correlated with one another (0.63 <|rs|< 0.80) [43].

The findings do not call into question the importance of multi-year sport-specific practice, childhood/adolescent performance development, and their support through TPP nurture. All the senior world-class athletes, senior national-class athletes, and high-performing junior athletes engaged in considerable sport-specific practice over multiple years [42, 43]; many had remarkable performance progress during junior age categories (achieving regional, national, and international junior championships [63]); and all were selected for a TPP at some age. However, a particularly accelerated childhood/adolescent development—typically via an early start, extensive main-sport practice, little or no other-sports practice, and early TPP involvement—is frequent among the highest-performing junior athletes but is rare among the highest-performing senior athletes ([42, 43], the present findings).

4.1 Theoretical Implications

The present meta-analysis complements a recent series of meta-analyses [42, 43, 63, 72, 73] that empirically tested the validity of the ‘hard core of assumptions’ [74] of traditional theories of giftedness and expertise [26, 75,76,77,78,79,80]. As mentioned above, the assumed premises (1) that achieving a high performance level in childhood/adolescence is a prerequisite for the long-term attainment of a high level of eventual senior performance, (2) that starting sport-specific practice at a younger age leads to a higher level of eventual senior performance, (3) that accumulating a larger amount of organized coach-led main-sport practice leads to a higher level of eventual senior performance, and (4) that accelerated childhood/adolescent performance progress leads to a higher level of eventual senior performance, have all been revealed to be at odds with the empirical evidence [42, 43, 63, 72, 73]. Likewise, the present meta-analysis empirically counters the assumption (5) that younger TPP involvement leads to a higher level of eventual senior performance. Taking the present study and recent evidence together [42, 43], predictors of rapid junior performance and of long-term senior performance are opposite in five aspects: starting age; amount of coach-led main-sport practice; amount of coach-led other-sports practice; age at commencement of TPP involvement; and age of ‘milestone’ achievement.

To illustrate for this discussion, typical performance trajectories of athletes who commence TPP involvement at younger versus older ages are schematically depicted in Fig. 2. Generally, TPPs include the selection of youth athletes and their ‘treatment’ by applying TPP measures to them. The present findings may thus be attributable to specific selection or intervention effects of TPPs, or an interplay of both. Many of the youth athletes selected at a young age have an early biological maturation (onset of puberty and growth spurt), are relatively old within their birth year, and have previously accumulated large amounts of specialized main-sport practice, with little or no other-sports practice [34,35,36,37,38,39,40,41,42,43]. When accelerated early performance progress rests on these factors, this early progress is often associated with reduced long-term sustainability, in that the performance trajectory subsequently flattens (Fig. 2) [34, 36, 42,43,44,45,46,47,48,49].

Fig. 2

Typical performance development of athletes who commence talent promotion program (TPP) involvement at younger versus older ages. Schematic illustration based on data of the present meta-analysis and [42, 43, 63]. The vertical Y-axis symbolizes the junior age limit. Early TPP entry is typically associated with early start to play one’s main sport, extensive specialized main-sport practice, little or no other-sports practice, and early achievement of performance ‘milestones.’ Late TPP entry is typically associated with later start to play one’s main sport, reduced main-sport practice, more other-sports practice, and older ‘milestone’ achievement

Once selected, the TPPs seek to further accelerate the youth athlete’s childhood/adolescent performance progress via expanded sport-specific practice and competitions, supported by corresponding environments, resources, and interventions [7, 9, 16, 22,23,24,25]. The strategy may further boost the youth athlete’s current rate of performance progress and lead to increased short-term performance during junior age categories (Fig. 2). However, it may further compromise long-term sustainability because the large amount of childhood/adolescent specialized practice leads to reduced efficiency of practice in subsequent years—the larger the amount of previously accumulated specialized practice the smaller generally the subsequent performance gain per added practice [42, 43]. In economic terms, these athletes’ diminishing return (Gossen’s first law: diminishing performance improvement per added practice amount [81]) is more pronounced than among peers who had only moderate childhood/adolescent main-sport practice combined with other-sports practice and competitions. The data of Barth et al. [43] have shown that senior world-class athletes began to have an advantage in efficiency of practice over senior national-class counterparts in late adolescence and this advantage in efficiency of practice peaked in athletes’ early 20s.

Furthermore, the expanded time demands, and cumulative physical load associated with early TPP involvement, impose additional opportunity costs on the youth athlete and increase their risks of later occurrences that hinder or end the athletic career (e.g., conflicting time demands from sport and school, declining academic performance, injury, burnout, dropout) [7, 9, 50,51,52,53,54,55,56,57,58,59,60,61, 82, 83].

In contrast, many eventual senior world-class athletes developed outside the TPPs, just based on the youth sport program of their home/local sport club or school, until older ages, thus remaining unaffected by potential dysfunctional effects of early TPP involvement. Senior world-class athletes’ childhood/adolescent investment pattern was typically characterized by reduced intensity (less main-sport practice), diversification of investments (two to three sports), greater personal, private contribution (later TPP support), longer total period of “product development” and longer deferral of reward (later ‘milestone’ achievement), reduced total childhood/adolescent costs and long-term risks (injury, burnout, dropout), yielding increased long-term benefit in terms of senior performance (see [42, 43, 84], which also include further in-depth discussion of explanatory hypotheses).

4.2 Practical Implications

Taken together, the recent and present evidence suggests several clear practical implications.

Given the negative correlation between early TPP involvement and long-term senior performance, the selection age of many TPPs may generally be postponed to older ages.

Just the youth athlete’s current performance level, be it performance in competitions or in standardized tests of physical and perceptual-motor performance, does not reflect their future potential and is therefore not a sensible selection criterion [16, 18, 34, 63, 72, 85, 86]. Above that, selecting by current performance may have a dysfunctional ‘radiating’ effect [31, 63], in that it stimulates all those seeking admission to TPPs, youth athletes, coaches, parents, and perhaps other stakeholders, to attempt to reinforce early acceleration of the youth athlete’s development prior to the selection age. Instead, provided long-term positive effects of a TPP have been proven, selectors should strive to identify and estimate indicators of the youth athlete’s potential for long-term development. In this context, a participation history of moderate main-sport practice while playing other sports over multiple years is a predictor of long-term potential and should be considered for talent selection, while also taking into account whether athletes were born earlier or later within their birth year and whether they have an earlier or later onset of puberty and growth spurt.

Relatedly, evaluations of TPPs, which often form the basis for their external funding, should not focus on participants’ current junior performance or their short-term progress, including participants’ progression to a subsequent TPP stage or level. This would elicit dysfunctional incentives in that it would further stimulate attempts to select the most advanced youth athletes and then further accelerate their current performance. Rather, evaluations of TPPs should consider their long-term sustainability and focus on the participants’ performance development through subsequent years into adulthood and the long-term senior performance level they eventually achieve, while considering their costs and risks.

Finally, TPPs should generally seek to enhance the youth athlete’s short-term and long-term benefits while limiting their costs and risks. Specifically, TPPs should limit additional opportunity costs in terms of time demands and risks in terms of amounts of practice and competitions and associated cumulative physical load imposed on the youth athlete.

4.3 Methodological Considerations and Future Directions

The study had several strengths, such as a large international sample including a broad range of individual and team sports; considering the most common types of TPPs; distinguishing short-term effects on early junior performance and long-term effects on later senior performance; comparing higher-performing and lower-performing athletes across regional, national, and international levels; comparing athletes within the same age category, type of sport, sex, country, and regarding age at entry into the same TPP, respectively; and a high quality of the primary studies. Nevertheless, several limitations should be acknowledged: (1) the correlational design of the primary studies does not allow for causal conclusions; (2) most athletes were male, from Olympic sports, and from Western industrialized countries (but not from the largest Western country, the USA). Effects of early TPP involvement may differ in Paralympic sports or in different sport systems such as the USA, developing countries, Eastern Europe, China, or Russia. (3) All athletes competed at a regional, national, or international level, which may imply restriction of range. Effects of early TPP involvement may differ among lower-level or more heterogeneous populations. (4) There were not enough data for moderator analyses across single sports. Effects may vary between single sports. (5) Successful senior athletes who were not selected for a TPP during junior age categories may not have been considered in primary studies. (6) The synthesized primary studies used linear methods of data analysis and did not consider non-linear analyses. Finally, although we used multiple databases, as in any systematic review, bias of availability, country, and language is possible.

A goal for future research is to investigate the extent to which opposite effects of the age at commencement of TPP involvement on short-term and long-term performance are due to specific selection or intervention effects of TPPs, or an interplay of both. That research should examine the effects of the individual and combined TPP measures applied to participants on their short-term and long-term performance, while also considering effects on other short-term and long-term outcomes such as athletes’ health, psychosocial well-being, academic achievements, and persistent sport engagement. Furthermore, researchers should seek to extend investigations into TPPs to populations that are under-represented in present research, especially female athletes, more sport-specific samples, in particular from Paralympic and non-Olympic sports, and samples from more countries, in particular the USA, developing countries, East European countries, China, and Russia. In this context, the present findings indicate that long-term TPP effects on eventual senior performance (and perhaps other long-term outcomes) cannot be inferred by extrapolating findings from junior samples [7, 19, 61, 87, 88]. Rather, relevant statements require comparison of higher-performing and lower-performing senior athletes regarding the TPP environments, resources, and intervention measures they were provided during childhood and adolescence.

It is also interesting to note that multiple decades of extensive research into talent identification (TID) [5, 15, 27, 29, 89,90,91,92,93,94,95,96,97,98,99] is contrasted by lacking research into the purpose TID is done for: i.e., the effects of the TPP interventions applied to the talent-identified TPP participants [16,