Researchers at Slovenia's Jožef Stefan Institute (IJS) and Ljubljana Faculty of Sport have developed a new method to predict adult height in children and teenagers, using extensive population data and artificial intelligence. The new predictions are much more precise than those based on methods currently in use.

Growth pattern of a child can often be inferred from their parents' growth curves, but usually these are not easily available. The most common approach used by paediatricians is therefore percentile charts, where the doctor checks whether the child has stayed in the same percentile rank throughout the years, IJS said.

This approach does not take into consideration the fact that adolescent growth spurt does not occur at the same time in everyone. Some adult height prediction methods involve invasive radiological examinations, and that their development was based only on a few hundred individuals.

But what Slovenian researchers did is create a new algorithm that is based on large-scale population data collected over decades of the national child fitness study.

The annual study SLOfit includes height and weight measurements taken from the age of 6 or 7 until the age of 18 or 19, and it has been carried out in Slovenia ever since the 1980s.

In order to predict future growth curve and adult height, the new algorithm, helped by AI, compares child growth curve with the height of individuals with the most similar characteristics in the database, which includes more than 16,000 primary and secondary school students.

The method is publicly available on the SLOfit website, and the research itself, which was done as part of the EU project Horizon 2020 CrowdHEALTH, has been published in the peer-reviewed journal PLOS One.

The study continues under another EU project, SmartCHANGE, which will focus on developing more effective predictions of chronic diseases, such as cardiovascular and metabolic diseases and different types of cancer, in children and adolescents in their adulthood.