To better understand and affect children's physical activity levels, recent research on the determinants of physical activity has expressed the need for studies that more precisely investigate the varying contexts in which parental activity is related to children's activity [9, 45]. The data presented in the current study add to the literature examining parent-child correlations of activity by examining gender-related differences as well as examining patterns during relevant weekly segments: weekend, weekday, and weekday after-school hours. The number of studies using accelerometer derived measures of physical activity to assess parent-child correlations in activity has been increasing, and this study is in line with this small, but growing, body of research.
In these data, we found that, overall, MVPA of parents and children were significantly correlated for many of the observed time segments (weekend, weekday, and weekday 3 to 7 pm). These associations remained significant in multivariate regression models. However, the association between parents' and children's sedentary activity was not as consistent. In the regression models, the only statistically significant finding was that fathers' sedentary activity was associated with children's on the weekend. We also found that children of two highly active parents engaged in more MVPA than children of parents who engaged in very little MVPA. The MVPA correlations stratified by gender of parent and child were interesting in that they appear to be gender specific (i.e., mothers' were correlated with daughters' and fathers' with sons'), especially for the weekend and weekday after-school time segments. Greater attention to gender specific association in future studies seems warranted.
Comparing our findings to those few existing studies of parent-child correlations using accelerometer derived measures of physical activity [12, 15–17] is difficult because; 1) the age ranges of children differ across studies, 2) some studies had data from only one parent, whereas others had data from both mothers and fathers, and 3) slightly different analytic methods have been employed across studies. However, in general, our findings comport with those that have found that parental physical activity is positively associated with an increase in children's physical activity  and those that find that children are more likely to be active when their parents are active [12, 17]. Notably, the one other study using accelerometer derived measures of physical activity found parents' sedentary activity was significantly correlated with their children's sedentary activity, but MVPA was not .
The study findings suggest several key points relevant to the ongoing efforts to better understand parent-child correlations in physical activity patterns. First, during times when most parents could potentially have the most direct influence on child activity (i.e., after-school and on weekends) both mothers' and fathers' MVPA were positively related to their children's. Although our data does not allow us to know whether parental modeling, support, shared activities, or combinations of such factors were responsible for the parent-child activity aggregation found, it seems clear that time periods outside of work or school are crucial targets for interventions that aim to involve parents.
Second, the results add to the increasing evidence regarding the importance of the after-school hours in youth behavior and health. Higher levels of physical activity have been found in the time period after-school in previous studies [31, 32, 34], and interventions that have targeted this time period have been successful in increasing physical activity and decreasing overweight in children and adolescents [46–48]. Previous studies have also reported distinct gender differences (boys > girls) in levels of MVPA during the after-school period that were not clearly replicated in our study [31, 34]. The boys in our sample had higher MVPA means during the after-school hours than the girls, but they were not statistically significantly higher (data not reported). Importantly, the MVPA levels of parents' were predictive of children's activity after-school, controlling for gender; although the correlations stratified by gender hint that these associations may be gender-specific. This deserves further study in other samples. Interestingly, in multivariate models we did not find that parents' sedentary time during the after-school segment predicted that of their children's.
Third, our study extends findings on the importance of having active parents as a predictor of physical activity in children. Although parental modeling has been well accepted as a possible mechanism for parent-child aggregation of physical activity, there have been very few studies that have looked at the impact of one versus two active parents using an accelerometer derived measure to quantify activity . Similar to findings by Moore et al. , we found a strong, linear relationship between the number of active parents and the activity levels of children. Children with two active parents engaged in greater MVPA than children where both parents were low in MVPA. We did not find any child gender effects, as Moore et al. did in their study, in which they found that parental activity was stronger for boys. In our study, in which the children were slightly older, children appeared to benefit substantially more from two active parents and this was true for both boys and girls. With respect to sedentary activity, children's sedentary activity time did increase in a graded fashion with the number of sedentary parents, but these were not statistically significant increases in mean sedentary activity time. In other words, high levels of sedentary time existed for all children, regardless of the sedentary classifications of the parents.
Our data, and the multiple analytical approaches we have used to interpret it, demonstrate the independence of MVPA and sedentary time  in parent-child correlations. Biddle and others have argued that even highly active people spend considerable time being sedentary , and this may have been the case among our participants. In sum, our data lend support to the notion that to increase childhood activity levels it may be fruitful to focus on improving the MVPA among the whole family, including both parents.
As in any study, our results should be considered with respect to the limitations. First, the sample size was small which may have had an impact on our ability to detect significant associations other than large effects. The sample was also fairly homogenous, especially in regard to parental education and weight status, which limits the degree to which our findings generalize to the broader population. Notably, however, the sample did include a larger percentage of non-white participants than is usually present in these types of studies. Population based studies of parent-child correlations using directly measured physical activity seem warranted. Another limitation is that the study design was cross-sectional. Longitudinal studies of parent-child physical activity correlations are needed especially in light of the well established finding that physical activity levels decline with age [24, 51–54]. It remains to be determined how parent-child correlations change through development. Another methodological factor to consider relates the use of accelerometers. Although we believe that accelerometer derived measures of physical activity are useful, their sole use can present some limitations. In general, accelerometers like the type used in this study, have been shown to be a valid method for assessing physical activity in children . However, these devices do not capture certain types of activity well, such as cycling, climbing stairs, or swimming . Further, future studies should be designed to provide more contextual information on what parents and children are doing during time periods of interest (e.g., time use data), which would increase understanding of the types of activities in which parents and children engage and if they are active or inactive together. Lastly, although not necessarily a limitation, the total amount of MVPA was higher in children than in parents. This is in line with the noted observation of declining physical activity with increasing age. This difference in the total amount likely reflects the different ways in which children and adults accumulate MVPA (e.g., spontaneous play vs. structured or planned activities). We were not able to capture that level of detailed information in this study. Future studies are needed to look at how parents and children of different ages accumulate their MVPA throughout the day and the mechanisms that help explain the parent-child correlations we observed in this study. Understanding these explanatory factors will be beneficial to interventions that aim to increase children's MVPA through increasing parental MVPA.