**3. Results**

Overall, 999 primary school children aged 7 to 9 years from a small rural community in the German state of Baden-Württemberg were included in the study (PreCOVID *n* = 801; COVID1: *n* = 91; COVID2: *n* = 107). There were no significant differences for age and BMI between the ten cohorts, but gender-specific differences in mean values in the test items.

Figure 1 shows the trends in test items for all measurement years overall and separately by gender (see Table A1).

There was a linear, consistent level of performance for the 20 m dash through 2017 in boys and girls. In 2018, an increase was observed for either gender, and this level of speed remained stable until the last measurement in 2021. The ANCOVA for the test item 20 m dash revealed that children in the cohort COVID1 were 0.20 s slower than in the PreCOVID cohort (F(1,859) = 15.89; *p* < 0.001; η<sup>p</sup> <sup>2</sup> = 0.018). The comparison with cohort COVID2 showed a significant difference of 0.23 s (F(1,861) = 18.69; *p* < 0.001; η<sup>p</sup> <sup>2</sup> = 0.021). The influence of the covariate gender was significant for both ANCOVAs (*p* < 0.001; *p* < 0.001).

The analysis of the test item balancing backwards showed an opposite gender-specific trend until 2015, followed by a peak in 2016 for both boys and girls. This increase stopped abruptly and tended to remain stable until 2019. No significant difference was found in the ANCOVA, with 0.22 steps between cohort PreCOVID and COVID1 (F(1,874) = 0.06; *p* = 0.813; η<sup>p</sup> <sup>2</sup> = 0.000). However, children in cohort COVID2 performed 2.03 steps better than cohort PreCOVID (F(1,893) = 5.67; *p* = 0.017; η<sup>p</sup> <sup>2</sup> = 0.006). The covariate gender had no significant influence on the analysis (*p* = 0.232; *p =* 0.215).

The mean values for jumping sideways revealed no trend. There were ups and downs through all cohorts, with a peak in 2020 and a minimum of performance in 2012 and 2017. There were inverse performance levels for boys and girls for the cohorts 2014 through 2016. Analyzing the differences for the test item jumping sideways showed that there was a difference of 4.32 fewer steps in PreCOVID compared with COVID1 (F(1,874) = 27.05; *p* < 0.001; η<sup>p</sup> <sup>2</sup> = 0.030). There was no significant difference found between PreCOVID and COVID2, with 1.24 fewer steps measured for PreCOVID (F(1,892) = 2.71; *p* = 0.100; ηp <sup>2</sup> = 0.003). The influence of the covariate gender was not significant (*p* = 0.423; *p* = 0.353).

A steadily declining trend was found for the test item stand and reach, with its minimum in 2018. The level of flexibility subsequently increased. This development was found for boys and girls equally, but with clear differences in the measured values. The ANCOVA revealed 0.88 cm more in COVID1 than in cohort PreCOVID, but the difference was not significant (F(1,871) = 1.44; *p* = 0.230; η<sup>p</sup> <sup>2</sup> = 0.002). PreCOVID had 1.00 cm less for stand and reach than COVID2, but this difference was also not significant (F(1,888) = 2.16; *p* = 0.142; η<sup>p</sup> <sup>2</sup> = 0.002). The covariate was statistically significant in both cohort comparisons (*p* < 0.001; *p* < 0.001).

Push-up performance was consistent over the cohorts before increasing in 2017 and peaking in 2019. A significant difference was found in COVID1 with 2.29 more performed push-ups compared with PreCOVID (F(1,875) = 28.63; *p* < 0.001; η<sup>p</sup> <sup>2</sup> = 0.032). There was a significant positive difference between PreCOVID and COVID2 of 0.94 (F(1,892) = 5.69; *p* = 0.017; η<sup>p</sup> <sup>2</sup> = 0.006). The covariate gender had no significant influence on either ANCOVA (*p* = 0.855; *p* = 0.924).

**Figure 1.** Trends for test items over each year of measurement overall, and for boys and girls separately.

The test item sit-ups improved consistently, but showed an apparent reversal in gender-specific performance for 2014 and the highest levels up to 2019. Analyzing cohort differences with the ANCOVA, the performance differed significantly with 1.95 more sit-

ups in COVID1 than in PreCOVID (F(1,863) = 9.87; *p* = 0.002; η<sup>p</sup> <sup>2</sup> = 0.011). In addition, with 1.36 more sit-ups, COVID2 was significantly better than PreCOVID (F(1,881) = 5.50; *p* = 0.019; η<sup>p</sup> <sup>2</sup> = 0.006). The covariate gender had a significant influence on the differences (*p* < 0.001; *p* < 0.001).

For standing long jump, initial measurements already showed significant differences between boys and girls. This difference was found for all cohorts with no apparent trend. However, the analysis revealed that children in COVID1 jumped 7.24 cm farther than children in PreCOVID (F(1,873) = 11.13; *p* < 0.001; η<sup>p</sup> <sup>2</sup> = 0.013). However, the comparison between PreCOVID and COVID2 showed no significant difference, with 0.56 cm more for COVID2 (F(1,890) = 0.08; *p* = 0.782; η<sup>p</sup> <sup>2</sup> = 0.000). The covariate gender was statistically significant for both cohort comparisons (*p* < 0.001; *p* < 0.001).

The mean values for 6 min run in the 2012 cohort differed significantly for boys and girls. Performance differed by gender over the measurement period, but the development of the overall sample revealed no trend. No significant differences were found between COVID1 and PreCOVID for the 6 min run, with COVID1 running only 15.31 m more than PreCOVID (F(1,857) = 0.91; *p* = 0.341; η<sup>p</sup> <sup>2</sup> = 0.001). The children in PreCOVID ran 16.94 m less than COVID2, but these differences were also not significant (F(1,866) = 1.23; *p* = 0.267; ηp <sup>2</sup> = 0.001). The covariate gender had a significant influence on the measured differences (*p* < 0.001; *p* < 0.001). See Tables A2 and A3 for the adjusted mean values of the ANCOVA in each test item and cohort.
