The ACDSi 2023/24 edition is the latest iteration of an ongoing, secular trend study, with its roots dating back to 1970–71. In those years, the Faculty of Sport at the University of Ljubljana launched a pioneer study, initially entitled “Comparisons of some motor and morphological parameters in primary schools in Slovenia.” This project has evolved considerably over the subsequent decades. Since its inception, the ACDSi study has been implemented in decennial cycles −1970/71 (40), 1983 (41), 1993/94 (42), 2003/04 (43), 2013/14 (38, 39), and 2023/24—making it one of the longest-running investigations of secular trends in child and adolescent physical fitness and its determinants worldwide.

The original 1970/71 study employed a nationally representative sample, selected through a multi-stage, stratified sampling design (40, 44). Ten research sites were chosen to represent diverse types of residence: from rural to rural-industrial, industrial-rural and industrial and various regions across Slovenia. At each study site, one primary school was selected, and beginning in 1993 an additional school in the capital region was included to ensure appropriate population representation. In 1994, the ACDSi's study scope expanded to include upper secondary school students, aged 15 to 18, across three novel sites. This expansion involved adding six upper secondary school centers, with different educational program specializations. Subsequently, each consequent edition has been spanning over two consecutive years: in the first year, primary school children are being assessed, followed by upper secondary school students in the second year. In each edition, the ACDSi study consistently includes approximately 5,500 children and adolescents aged 7 to 18 years, representing around 2% of the entire population within the age range (Table 1). Notably, reforms in Slovenia's education system introduced an earlier primary school entry age for the 2003 cohort, expanding the sampling frame to include children as young as six. In earlier editions, each age and sex cohort included around 200 participants.

The scope of the ACDSi study has evolved over time. Initially, the focus was on physical fitness and physical anthropology, employing standardized tests prevalent in the former Yugoslavia and Eastern European countries (44–46). In 1993, the inclusion of tests from the Eurofit test battery (47) were incorporated, broadening the study's scope to encompass a more internationally comparable range of fitness tests.

Since its inception, the ACDSi study has progressively expanded its analytical scope to investigate key correlates and determinants of physical fitness and their interrelationships, including somatic development, motor abilities, cognitive functioning, social factors, and environmental influences. This expanded focus allowed for a more comprehensive understanding of physical fitness and factors shaping youth development in any given generation.

Although some data collection methods have varied somewhat across different study editions, the core dataset remained consistent, comprising of:

Unfortunately, the original 1970/71 individual-level data was lost in a fire, limiting comparisons to summary data published in the 1970/71 study report. Table 2 outlines the specific category of variables from each edition, including the updated 2023/24 dataset. Supplementary Table 1 presents detailed dataset information from all editions.

The Faculty of Sport at the University of Ljubljana leads the administration and implementation of the ACDSi study. The Principal Investigators oversee all phases of the research process, including study design, funding acquisition, coordination with local teams and participating schools, fieldwork organization, training of the measurement team, and data management—encompassing cleaning, analysis, and interpretation.

Preparations for each study edition start at least 1 year in advance. We start by informing the municipality mayors and school principals to ensure their support early on. We then collaborate with physical education teachers in each jurisdiction to appoint school-level coordinators.

These local coordinators work with our core research team to communicate with other teachers and parents, integrate the study into the school's annual teaching plan, collect parental consent and child assent, and coordinate the measurement schedule. Based on completed consent forms, the core research team creates a detailed participant list, each with unique IDs.

To ensure accurate measurements, the core research team recruits and thoroughly trains student investigators in June and September of each study year. During each testing session, a team of between 20 and 25 student investigators and experienced researchers is deployed to a given location. The team is then split into four groups, each comprised of a group lead. The four groups and core responsibilities are:

The described rigorous organizational procedures have been presented across all data collection phases and ensure consistent and reliable data collection between each of the ACDSi study editions.

The primary objective of the 2023/24 ACDSi study edition was to assess the current state of somatic and motor development in Slovenian children and adolescents to compare it with data from previous editions of ACDSi study. This involved addressing the following specific research questions:

What are the current levels of cardiorespiratory fitness, muscular fitness, body composition, and anthropometric characteristics in Slovenian children and adolescents, and how do these levels vary by age and sex?

A list of more detailed descriptive research goals included:

What are the socioeconomic characteristics of Slovenian children and adolescents, and how do these characteristics relate to their health, movement behaviors, and physical fitness?

In addition to descriptive research goals, we also established several prediction goals based on existing literature. These were to determine or observe:

How do socioeconomic characteristics and home-environment attributes relate to adolescents' cardiorespiratory fitness, muscular fitness, body composition, overweight/obesity, and 24-h movement behaviors?

The ACDSi is a decennial, cross-sectional sentinel-site study conducted since 1970–71, involving 11 primary schools and 10 upper secondary schools. The 2023/24 edition aimed to recruit a total of n = 4,200 participants, with a target of n = 150 per age-by-sex group (6–18 years). This corresponds to approximately n = 2,700 primary school students (6–14 years) and n = 1,500 upper secondary school students (15–18 years). This target sample size is lower than in previous editions (n = 200 per group), reflecting demographic changes—specifically, the decreasing number of school-aged children in Slovenia—and optimization of study costs. Importantly, power analyses confirmed that the revised sample remains sufficient for all planned statistical comparisons.

No incentives were provided to participants for their participation.

The ACDSi 2023/24 study employed a two-stage cluster sampling design with schools as primary sampling units and classes as secondary ones. Sample stratification at school level included:

The target sample size was 300 participants per grade, corresponding to 2,700 students across nine primary school grades and 1,200 students across four secondary school grades. The actual sample size (2,721 in primary schools and 1,139 in secondary schools; Table 1) closely matched the planned targets.

By following this sampling strategy, the ACDSi 2023/24 study aimed to obtain a representative sample of Slovenian youth consistent with previous cycles (Figure 1).

The initial target for the ACDSi 2023/24 study was deliberately inflated to n = 6,056 participants (3,151 primary school students and 2,905 upper secondary school students) to account for the anticipated non-response and dropout. This upward adjustment was based on historical rates of approximately 20% for primary school and 50% for upper secondary school age groups. The final sample size of participants who successfully completed all fitness testing, anthropometry examinations, and questionnaires was n = 3,853 (Table 3; five participants did not provide information on sex and birth date). This corresponds to 63.7% of the initially invited sample and represents 1.35% of the total Slovenian population aged 6–18 years in 2024 (48).

In addition to the school-based measurements, paper questionnaires were distributed to parents of primary school students. Considerable effort was devoted to encouraging active participation, which yielded a satisfactory response rate of 89.6 % of obtained questionnaires.

All tests and measurement protocols used in the ACDSi 2023/24 study have been validated for use in children and adolescents within the targeted age ranges (see Table 4).

Fitness testing, anthropometric measurements, and questionnaire administration were conducted between September 1 and October 4, 2023, for primary school students, and between September 12 and 27, 2024, for upper secondary students. Measurements were organized between 8:00 AM and 2:00 PM and typically completed within 1 or 2 days per school, depending on the sample size. Every child or adolescent was present for measurements in 1 day.

A trained measurement team composed of student investigators from several faculties of the University of Ljubljana (e.g., Sport, Biotechnical Sciences, Health Sciences, Arts, and Education) conducted all assessments. Three senior researchers acted as field supervisors overseeing the procedures for anthropometry, fitness testing, and questionnaire administration. Any equipment not typically available in school gymnasiums was provided on-site by the core research team. All testing equipment was routinely calibrated every morning before testing began, and throughout the testing period (when necessary). A complete list of the ACDSi 2023/24 research team is acknowledged in the Supplementary material.

Anthropometric measurements were conducted in smaller gyms or classrooms maintained at a temperature of 20 °C−24 °C. Fitness tests were administered in school gyms under similar temperature conditions. Outdoor running tests were conducted only when the temperature exceeded 10 °C, and in the absence of heavy rain or wind. Web-based questionnaires were completed by participants in school computer labs, before or after fitness and anthropometric measurements in the gym, with groups of up to 22 students per session.

Each gym measurement session involved 40–60 participants, divided into two groups: one for anthropometric measurements, and the other for fitness tests. Each session lasted approximately 90 min. Blood pressure was measured for all participants in seated position before any physical exertion took place. A standardized warm-up was led by the research team immediately preceding fitness testing.

All participants aged 11 and older completed a web-based questionnaire in a computer room under the guidance of a responsible investigator. Upon arrival in the computer room, each participant received a unique login ID.

After logging in, students completed the questionnaire, while investigators provided individual assistance when required.

After completion of the on-site student measurements, questionnaires for parents of children aged 6–14 years were distributed. These were to be completed at home and returned to the school coordinators the following day.

All data collected in the ACDSi 2023/24 study are jointly owned by the research team, with the principal investigators responsible for database management. The database, together with detailed measurement protocols, is available to the scientific community upon request for comparative analyses.

A detailed description of all anthropometric measures and fitness test items can be found on the study website in the document Motor_anthropometry_tests.pdf.

All hard copy data were subject to double manual entry by the same person into a secure database to minimize transcription errors. Following the entry, descriptive statistics were computed for all variables and checked against previous ACDSi study editions to potential deviations and extreme values. Outlier detection was performed using a standardized, statistical method. In detail, the 10 most extreme outliers were manually checked. These were identified based on the largest absolute values of Studentized residuals derived from linear model that included sex, age, sex^*age interaction, and the additional variable, which had the highest Pearson correlation within the predicted variable. Possible additional outliers (i.e., noise) and smaller outlier groups were detected using the DBSCAN clustering technique, described elsewhere (67).

Missing data handling will depend on the specific analysis. We will preferentially select analysis methods that generally allow model building in the presence of missing data. If this will not be possible, and if the proportion of missing values will be small (< 5%), we will include only participants with complete data. For a higher proportion of missing data, particularly if a sensitivity analysis will indicate a risk of bias from non-imputed data, we will perform imputation using the random forest method.

In fitness tests results, the highest performance score will be retained for analysis, whereas in anthropometric measurements, the median of repeated values will be used to ensure accuracy and reduce measurement error. Indexes such as BMI, sum of skin folds, and other relevant metrics will be calculated according to established methods. Additionally, tests which have summative scores—such as components of motivation, self-concept, and somatotype—will be derived using their predefined protocols. The socioeconomic environment will be assessed using the FAS (68).

Given the school-based design of the ACDSi study, clustering of results within schools is possible and should be considered in data interpretation. This means that differences observed between children of a given age or constitution might be more pronounced between schools than within individual schools. To account for this potential bias in the dataset, mixed-effect models with school-effect as random factor will be employed to address this clustering effect.

In addition to descriptive statistics and bivariate analyses, which will be used to characterize data distribution and identify basic associations between predictors and outcomes, multilevel multivariate analyses may be performed depending on the specific research questions addressed.

For establishing longitudinal trends Multilevel/Mixed-effects models will be used, with time (wave of a study) as a fixed effect, school clustering as a random effect (schools nested within time points) and individual-level characteristics (e.g. age, sex) as covariates. As a secondary analysis, we plan to use Age-Period-Cohort (APC) analysis to better understand the mechanism of change (cohort vs. period) and age effect.

Descriptive statistics will be stratified by sex and age. When testing hypotheses and constructing statistical models, analyses will generally begin with a base model for each outcome that includes sex and age (and other relevant stratification variables) as covariates. If model assumptions are violated—for example, unequal variances (heteroscedasticity) or substantially different subgroup structures—we will perform separate analyses for each sex and/or age group.