The validity of automatic methods for estimating maturation stage in young athletes: A comparison of the Maturo smartphone application and sport science expert evaluations

Abstract

Background: Biological maturation refers to the progressive process through which individuals transition toward an adult state during growth and development. To address the challenges posed by differences in biological maturity and the limitations of existing testing methods, particularly in adolescent sports contexts, there is a pressing need for a non-invasive method that is convenient, accurate, and broadly applicable to monitor the biological maturity of adolescent athletes comprehensively. In response to this need, a maturity assessment method based on the smartphone application Maturo has been developed. This study evaluates the accuracy and validity of the Maturo software, an automated tool for estimating biological age and related maturation metrics.

Methods: A sample of 103 actively training teenage athletes aged 9 - 17 years. The sample included 76 males (age = 11.74 ± 1.55 years, mean ± SD) and 27 females (age = 13.95 ± 1.40 years), all without medical conditions that might impact growth or development.

Results: Compared to traditional expert evaluations, the intraclass correlation coefficients (ICCs) and Pearson correlation coefficients demonstrated reliable positive correlations and significant agreement between the Maturo software and expert methods across multiple metrics, such as biological age (ICC = 0.965, R = 0.97), corrected biological age (ICC = 0.973, R = 0.99), predicted adult height (ICC = 0.991, R = 0.99), and percentage of adult height achieved (ICC = 0.955, R = 0.97). The Bland-Altman plots provided additional evidence of the validity of the Maturo software estimations, showing low systematic error in most measures. The linear regression analysis produced excellent adjusted R² values: 0.95 for biological age and 0.99 for anticipated adult height. The Maturo approach demonstrated a high level of dependability in classifying teenagers into groups based on their maturity status and timing. The κ coefficients of 0.93 for maturity status and 0.82 for maturity timing indicate a nearly perfect agreement with the expert technique.

Conclusion: While the Maturo software's non-invasive nature, cost-effectiveness, and ease of use could make it a potential tool for regular monitoring of growth and maturation in young athletes, its promising results in assessing maturation should be interpreted with caution due to limitations such as sample size and demographic constraints. Further longitude research with larger and more diverse populations is needed to validate these preliminary findings and strengthen the evidence for its broader applicability.

Keywords: Anthropometry; Biological maturation estimation; Maturation assessment; Maturo software; Youth athletes.

Graphical abstract

Fig. 1

Flowchart of the Maturo software process. AI = artificial intelligent.

Fig. 2

Data collection for the human expert method.

Fig. 3

Data collection for the Maturo method.