The goal of our study was to evaluate differences in motor skills and in physical activity according to weight or gender in 2- to 4-year-old children attending child care center. No significant weight status- or gender-related differences in overall motor skills were found. However, total physical activity and time spent in MVPA and VPA were higher while sedentary activity was lower in the 2- to 4-year-old boys compared to girls.
To our knowledge, weight status-related differences have not been studied in children that young. A few studies in 4- to 6-year-old preschoolers examined differences in motor skills according to weight status [19–21, 36], and their results were controversial. One study found no differences , while several other studies [19–21] found globally better motor skills (particularly in the more dynamic tests) in healthy weight compared to overweight children. The reasons for the discordance between studies are not clear as many of the tests applied (locomotion, agility and balance) were quite similar. The authors hypothesized that the differences might be due to the fact that performance differences may become apparent at a later age and subsequently increase with age [21, 36]. We would need more studies to investigate this hypothesis, as it is refuted by a study in school children that did not find any weight-related differences in motor skills . However, our data, generated in a relatively large population of over 500 2-to 4-year-old children, would be in accordance with this assumption: We found a trend for a lower motor performance in overweight children, but no significant weight status-related differences in motor skills. Yet, the few number of overweight children might have reduced our power. Using weight status as a categorical variable is a crude and conservative way to test for the effect of weight on motor performance. However, using BMI instead of weight status did not change the results in a relevant way. The lack of large significant differences in motor skills in this age group highlights the importance of an early prevention to reduce the barriers of later overweight related to physical fitness. Thereby, the child care center may represent an ideal setting.
No weight status-related differences in total physical activity, MVPA, VPA or sedentary activity were found. In school children, differences in physical activity between healthy weight and overweight children are well documented [11, 12]. Even in a defined setting during school time, 8- to 10-year-old healthy weight children spent more time in MVPA compared to overweight children . Data in preschool children are more controversial [22, 23]. In a small study (n = 56), Metallinos and al.  observed that overweight 3- to 5-year-old children spent significantly less time in VPA than their healthy weight counterparts, but total physical activity (counts per minute) did not differ. In another study investigating 3-to 5-year-old children , significant weight status-related differences in time spent in MVPA and VPA were found in boys (n = 118), but not in girls (n = 127). Again, total physical activity did not differ. Niederer et al.  studied 4- to 6-year-old children (n = 613) and found no differences in total and VPA between healthy weight and overweight children, but significant differences in these measures between healthy weight and obese children. Differences between studies could be explained by age or methodological differences such as cut-offs to define overweight/obesity [28, 37, 38] or the different physical activity intensities [25, 39, 40]. The relatively low PA cut-offs compared to cut-offs  used in other studies may also have contributed to the absence of significant differences. Previous reports [21–23] did not compare sedentary activity between healthy weight and overweight preschoolers, although sedentary activity is an important measure in regards to overweight and cardiovascular risk . In the current study, no differences were found between healthy weight and overweight children regarding sedentary activity. Overall, our results and the results of previous studies suggest that weight status-related differences in motor skills and in physical activity start to develop around the preschool age and these data underline the importance of early preventive activities.
In the current study, boys performed better than girls in two motor skills (running and climbing stairs), but these differences were no longer significant after summing up the motor skills into the Basic Motor Score. The difference in these two motor skills did also not remain significant after adjusting for differences in physical activity. As we noticed a gender difference in physical activity, these results might imply that girls' motor skills could have improved if they were getting the same amount of physical activity as boys. Interestingly, a study in preschoolers  observed that girls in the highest motor skill tertile spent significantly less time in VPA than boys in the highest tertile. Thus, the relationship between PA and motor skills seems to be complex and may also depend on the intensity of physical activity and the level of motor skill performance. Our results regarding motor performance are to some extent in accordance with the literature. Although Cliff and al.  observed in a small study (n = 46, 4.3 ± 0.7 years) that girls performed better than boys, Fisher and al.  found no gender differences in a study including 394 children (age 4.2 ± 0.5 years). Differences in sample size and the choice of motor tasks might explain these contradictory findings. Although the motor test used by Fisher and al.  (Movement Assessment Battery)  is quite similar to the one used by Cliff and al.  (Test of Gross Motor Development) , some skills were evaluated in one study but not in the other (balance and skips in the Fisher and al. ; run, gallop, leap, slide and ball dribble in the Cliff and al. ). Taking the two larger studies together (Fisher et al. and the current one), there seems to be no significant gender differences in global gross motor skills in young children. However, it would be interesting to investigate different gross motor skills domains (i.e. locomotive, manipulative or balancing tasks) in order to explore more precisely potential gender differences, as suggested previously .
Contrary to motor skills, we found significant gender differences in total physical activity as well as in the time spent in MVPA, VPA and in sedentary activity. Interestingly, larger differences were observed in the more intense physical activities. Although we only measured over a 8 hour period during one child care day (mean wearing time of 6 hours), our results show that even in a defined setting like child care center, with the same space and a globally similar timetable, boys are more active and less sedentary than girls. This finding is in accordance with previous studies in older children [22, 26]. Similarly, data in infants, albeit not with accelerometry, confirm our data . Our observed gender differences were not modulated by age and thus were equally present in 2- to 3-year-old children compared to their 4 year-old counterparts. As differences are observed at any age throughout lifespan [22, 24, 26, 46, 47], one wonders if gender differences in physical activity are innate or culturally-driven (i.e. impact of "nature versus nurture"). Regardless, child care center educators should be sensitized to these found differences in order to minimize gender inequities in physical activity opportunities in young children.
The present study has a number of limitations. The child care-based investigation of physical activity was performed during one single day (mean wearing time of 6 hours) and only during child care which may not be representative for the whole day, or be representative of a child care day. Considering the literature a minimum of three days would have been more valuable for this assessment . However, Trost and al. suggest that the variability of this assessment observed over a single day seems to decrease as children get younger. Measurement restricted to the child care center setting can also be an advantage allowing detecting differences in a defined setting. But we cannot differentiate whether the social and physical environment in the respective child care center was more attractive to boys than for girls or whether boys may be genetically more active than girls. The motor skills assessment for this specific age has been validated . In our own pilot study, test-retest and inter-rater correlations were 0.5 and 0.7, respectively. Compared to other motor tests [43–45], reliability and validity of the motor skills tests are only moderate. However, this test has the advantage to assess motor skills performance in a relatively short time, as six children can be assessed within 15 minutes if two raters participate.
Strengths of the study are its large sample size, the investigation of motor skills and of child care-based objectively measured physical activity and the young age of the children.