This study compares the skeletal maturation of a current Central European healthy cohort with that of the Rotterdam cohort, a historical, healthy collective from 25 years ago in a comparable geographical location. One strength of the current study is the large number of cases analyzed, with an average of 83 patients per year of life. Determination of BA by AI not only is economical in terms of personnel resources when dealing with such a high volume of cases but also exhibits a lower precision error than that observed in the case of human readers [2, 11,12,13,14,15,16]. While the deviation of BA from CA was negligible in younger children, boys under 8 years of age and girls under 11 years of age, a clear acceleration of BA by up to 0.6 years in boys and 0.9 years in girls was observed in older children (boys above 8 years of age and girls above 11 years of age). A misjudgment of BA by the AI software is very unlikely, given its repeatedly documented reliability and high concordance with the estimation of BA by another AI-supported program. Several studies have examined whether the BA according to G&P corresponds to the CA of a current Caucasian population [17]. A meta-analysis in more recent populations (up to a decade ago) found that the G&P atlas, although of limited accuracy in Asian and African populations, is still reliable in Caucasians today [17]. Nevertheless, an acceleration of BA was reported 40 years ago, especially in adolescents during and after puberty, though on a lower level compared to that in our study (0.2 years in boys, 0.13 years in girls) [18]. A more recent study conducted in Germany found larger differences between BA and CA, by 0.49 years in boys and 0.39 years in girls, which is more similar to our results [19]. It appears that the overestimation of BA mainly occurs in older children, whereas BA tends to be underestimated in younger patients [6, 20]. However, in the cohorts examined in the studies mentioned before, no significant deviations were found between BA from CA. It is, therefore, more interesting to note that the Leipzig collective differs significantly in terms of maturation from a geographically nearby (being 700 km distance away from the current cohort), healthy, 25-year-old collective that was examined using identical AI software [6]. The differences were statistically significant from late childhood onwards (boys from 8 years of age, girls from 11 years of age) and amounted to up to 0.76 years for boys and 0.74 years for girls. In their “century-long study,” Boeyer et al. described an objectively earlier onset and complete ossification of the epiphyseal joints in children born in 1995 compared to those born in 1935 [3]. Possible systemic factors are described (e.g., an increase in body mass index with changed dietary behavior) as changing socioeconomic conditions, which influence sex hormone levels in adipose tissue even before the perceptible onset of puberty. In this context, a trend towards an increase in body mass index and an earlier onset of puberty and menarche was described [21,22,23,24].

Other articles, which analyzed a current cohort using BoneXpert, also describe a trend towards a deviation of BA from CA similar to that of the current study, especially from puberty onwards, whereas no significant deviations were seen in the age group under 10 years [25]. In a Mexican population, boys aged 14–16 years and girls aged 12–14 years showed a BA about 1 year older than the CA [23].

This study has several limitations: The most important is that the retrospectively recruited collective is not a proven healthy representative collective, but one of children who had received a hand radiograph for trauma (excluding a trauma sequence). With a certain statistical probability, children with growth disorders (accelerated as well as retarded) can also be found among them. If it is assumed that the proportion of growth-retarded children is neither higher nor lower in children with trauma to the hand than in the overall population, these patients should account for only about 5% of the cohort and not significantly affect the mean, as delay and acceleration occur equally. It is also unlikely that certain ethnic groups suffer hand trauma more frequently.

In summary, the BA of the current Central European population and that of a two-decade-old population (Rotterdam cohort) differ. Whether this is due to the geographical distance between the base cities of the two cohorts or accelerated bone maturation in the last two decades must be clarified by further studies.