Parent-child interactions and objectively measured child physical activity: a cross-sectional study

Increasing the proportion of children who engage in regular physical activity continues to be a public health priority. Physical activity provides important health benefits for children including increased physical fitness, reduced body fatness, favorable cardiovascular and metabolic disease risk profiles, enhanced bone health, and reduced symptoms of depression and anxiety [1]. Despite the known benefits, few US children meet the current recommendation of 60 minutes of physical activity per day [2], and unless they participate in a structured sport children are likely to become more sedentary as they age [1, 3, 4]. The problem of physical inactivity may be even more severe in rural areas of the US. Rural children often do not participate in after school sports due in part to limited opportunities and transportation barriers [5]. This may contribute to a greater prevalence of leisure-time inactivity and obesity among rural residents as compared to their metropolitan and suburban counterparts [3, 4].

Understanding the factors that influence child physical activity habits is important if we are to try to close the gap between current health behaviors and national recommendations. Health-related behaviors and patterns are established during childhood and adolescence, and evolve within the context of the family [6]. Although parents represent only one possible area of influence (e.g. peers, school), socialization of many behaviors occurs within the family, with parents' beliefs, attitudes, and behaviors substantially affecting children's health [7].

Parents, in particular, exert their influence on their children directly through specific parenting practices and indirectly through their parenting style [8]. Parenting practices are the behavioral strategies that parents employ to socialize their children. In the context of physical activity, practices such as logistical and emotional support and direct modeling have been shown to be positively associated with higher child physical activity levels [9–13]. Other parenting strategies that may positively impact children's physical activity include incorporating activity into family recreational routines, making activity-related equipment available at home, identifying safe places in the community that children can easily access, and finding activities to do outdoors for all weather conditions [14].

In 2000, Sallis et al. concluded that parental influence over child physical activity was indeterminate [15]. A more recent review by van der Horst and colleagues [16] found that several factors were positively associated with child physical activity including gender (male), self-efficacy, parental physical activity (for boys), and parent support. In a meta-analytic review, Pugliese and Tinsley [7] also found that a moderate positive relationship exists between parental support and modeling behavior and child physical activity levels. In fact, children had a relative risk of being inactive that was 1.41 times greater if parents did not engage in certain socialization behaviors (encouragement, instrumental, and modeling behaviors) than when they did engage in those behaviors. Other parental behaviors may also influence child physical activity. In a study of 800 Latino parents and their children, Arredondo et al. [17] found that parental reinforcement and monitoring were both positively associated with child physical activity. However, the authors noted that, in general, these broader aspects of parenting behaviors toward child physical activity remain understudied, perhaps due to a paucity of measurement tools to assess these constructs [18].

Certain parental practices may be especially important for particular groups of children. Compared to urban areas, rural communities face higher obesity rates for children, and lower physical activity levels may contribute to this difference. Oleson and colleagues [19] found that the distance between homes and physical amenities is great in rural areas, which suggests that children are dependent on their parent (or caregiver) for transportation to places to be physically active. Rural areas also tend to lack sidewalks and other amenities that would support recreational or leisure time activity [20]. In general, there are few rural-based studies in youth suggesting that this population requires more attention and study.

It is also important to note that parenting practices are related to, but distinct from parenting style. According to Darling and Steinberg's conceptual model [21], parenting style is theoretically independent of specific socialization content and influences child development indirectly by changing the effectiveness of the parenting practice. It is based on the idea that parents' attitudes and the beliefs they hold about how they should rear their children result in a two-way interaction that defines the emotional climate of the parenting environment. This dynamic process alters how children view their parents and thus changes how receptive children are to their parents' socialization demands. In other words, parenting style can either undermine or facilitate the parenting practices a parent employs to socialize his or her child.

This model is based on previous work by Baumrind [22] who defined parenting style using two dimensions of parental behavior: responsiveness or nurturance to and demandingness or control of the child. Responsiveness/nurturance was defined as "the extent to which parents foster individuality and self-assertion by being attuned, supportive, and acquiescent to children's requests" while demandingness/control refers to "claims that parents make on children to become integrated into society through behavior regulation, direct confrontation, maturity demands, and supervision of children's activities." A parent with an authoritarian style (high demandingness, low responsiveness) attempts to control their child's behavior with little regard for the child's needs and strict obedience to the parent while an authoritative parent (high demandingness, high responsiveness) provides encouragement for the child to express independence, clear set of boundaries, and open communication. A permissive parenting style (low demandingness, high responsiveness) expresses greater parental acceptance toward the child, but provides few boundaries and places few demands on the child and, lastly, an uninvolved parenting style (low demandingness, low responsiveness) offers few parental boundaries and little interaction with the child.

In general, evidence supports the association between authoritative parenting and positive child health outcomes across multiple domains [23–25]. Within the realm of physical activity, few studies have examined the role of parenting style in shaping children's activity-related behaviors. One study by Schmitz et al. [26] found that authoritative parents have children who report higher levels of physical activity. A more recent longitudinal study of parental influences on adolescent physical activity showed that factors such as family cohesion, parent-child communication, and parental engagement positively predicted moderate-to-vigorous physical activity for boys and girls after one year [27]. The authors suggest that these parenting behaviors may be reflective of an authoritative parenting style in that they provide support for adolescent physical activity and are balanced by appropriate levels of autonomy. Yet, the majority of studies looking at parental influence on child physical activity have either not measured parenting style or not examined this construct in a systematic way. Knowing that parenting style has an impact on multiple domains and appears to be related to child physical activity, more work is needed in this area to better understand the parent-child relationship in the context of physical activity.

To fill this research gap, the current study explores the relationship between parenting styles and practices, and children's physical activity levels. The first aim was to understand the association between general parenting styles and activity-related practices of rural US parents. It has been argued that if practices are independent of a parent's style, then certain practices could be targeted singly while those that are linked require interventions that treat underlying family dynamics as a whole [28]. The second aim was to evaluate the relationship between parenting style, parenting practices, and child activity while controlling for known covariates. A weakness noted in the literature has been the reliance on self-reported activity; therefore, we used an objective measure - accelerometry - to capture children's activity level. In addition, we examined other variables including child weight, which may be related to children's physical activity. Finally, following the theoretical framework described by Darling and Steinberg [21], we tested the moderating role of parenting style on the relationship between activity-related practices and children's physical activity (Figure 1).

This paper describes the parenting styles and activity-related practices of a group of US families living in under-resourced rural communities. One of the primary goals of this study was to assess whether parenting styles and practices were associated with child physical activity. Our findings suggest that both parenting style and certain parenting practices are significantly associated with child physical activity. This is consistent with the limited amount of research in this area [9, 17, 26, 50] and also extends this work by documenting other aspects of parenting style (permissive and uninvolved) that have not been previously investigated with respect to child physical activity. In this study, permissive parenting style was associated with the most minutes of child physical activity and uninvolved parenting style the least. An authoritative parenting style was not associated with child physical activity in this study. Our findings may be in contrast to other studies [26] because other studies have not measured parenting style in a comprehensive or systematic way (e.g. studies that have focused solely on authoritative/authoritarian typologies). Given that rural children living in under-resourced communities have limited access to physical activity opportunities, parents may need to provide a high level of support and encouragement for their child to be active and also place few restrictions or demands on their children's time (characteristic behaviors of a permissive parenting style). This may also be supported by the finding that providing children with logistical support also resulted in more minutes of MVPA for children.

To our knowledge, no study has examined the relationship between specific activity-related parenting practices and parenting style. Similar work in other disciplines has demonstrated a link between feeding practices and parenting style [28, 46]. For instance, working with a different sample of rural parent-child dyads, Hubbs-Tait et al., [28] found that restriction and pressure to eat were negatively correlated to authoritarian and permissive parenting styles. However, Hughes and colleagues [46] found that authoritarian parents were more likely to put pressure on their children to eat, permissive parents were less likely to use restriction, and authoritative parents were more likely to monitor. We were surprised to find few relationships between parent's activity-related practices and their general parenting style. As expected, praising a child for his/her physical activity habits (reinforcement) was positively associated with an authoritative style, but contrary to our expectations, both discipline and monitoring of child physical activity were negatively associated with an authoritarian parenting style.

There may be several potential reasons why we found few relationships between parenting style and activity-related parenting practices. First, the sample size was small. Although sufficient to run these analyses, the loss of accelerometer data was slightly higher than expected and additional studies with larger sample sizes are warranted. Second, these findings suggest that more work is needed to understand the goals and values parents place on their child's physical activity that may differ from their general views of parenting. For instance, a recent study found that parents of 5-7 year old children in Spain valued socially acceptable behaviour over their child being active, which led them to have a more relaxed attitude toward their child's activity [51]. Lastly, the control, reinforcement, discipline and monitoring subscales were measured by one or two items and may therefore not adequately represent the construct they attempt to measure. Although the authors of the PEAS instrument contend that previous research has found that two items may sufficiently identify a factor, future research should develop additional items to assess these constructs as they relate to child physical activity habits [18].

It is also unclear why parenting practices other than logistical support were not independently associated with their child's physical activity. In a longitudinal study of adolescents, Ornelas and colleagues [27] did not find an association between parental monitoring and physical activity despite their hypothesis that this type of parenting strategy -if either too directive or restrictive - may negatively impact children's physical activity. As mentioned previously, more work is also needed to adequately measure the constructs related to parental discipline, control, monitoring, and reinforcement of their child's physical activity habits. However, findings from this study suggest that parenting style may in fact moderate the relationship between parent's activity related practices and their child's physical activity. For instance, monitoring and reinforcement were only associated with child physical activity when expressed in the context of a permissive parenting style. This provides some validation to the rationalization by Ornelas [27] that certain practices may be beneficial, but only when the emotional context within which they are expressed to the child is balanced appropriately. Overt support and encouragement of healthy behaviors is likely to be important in children's adoption of health behaviors such as physical activity, however, parents need to be aware of the emotional context employed. Davison and Campbell [14] theorize that children are responsive to parents' moral support of their physical activity (e.g. watching them compete in athletic events) and the general feeling that physical activity is valued in the family, yet may resent physical activity and consider it a chore when it is regimented or promoted as a method of weight loss. This may be one possible explanation for the finding in this study that the relationship between children's accumulation of total counts of activity and parental monitoring was negative when in the presence of an unpermissive parenting style.

Additionally, it has been suggested that parents may take a different approach to their parenting strategies toward physical activity among children who are overweight than toward those who are not [52]. In this study, child BMI z score was not related to any physical activity variable. Since the majority of children's active time was spent in moderate, not vigorous activity, further investigation is warranted regarding the intensity of rural children's physical activity and parent's perception of their child's activity level. A large proportion of children in this sample were either overweight or obese, which is consistent with other literature suggesting that for rural children, overweight prevalence is higher than in urban areas [4, 53, 54]. These findings may suggest that the level of energy expenditure by children in this study has been insufficient to prevent excess weight gain. More work is needed to understand why children living in rural, under-resourced areas do not engage in more vigorous activity, especially if this level of activity is necessary to address the growing prevalence of overweight among children [55–57].

It is important to evaluate these findings in the context of the study limitations. The small sample size would not support certain subgroup analyses (e.g. separating mothers vs. fathers or girls vs. boys). The sample also represents a diverse group of parent-child dyads, yet generalizability may still be limited given that the sample is predominantly mother-daughter dyads, the focus is on families living in under-resourced rural areas, a lack of representation from other minority groups, and the randomly selected, but convenient sample, from which the participants were recruited. Findings may be specific to rural areas, which despite its characteristic open space often has fewer resources for children to be active as compared to urban environments [58, 59]. Since this was a cross-sectional study, one cannot discriminate the direction of influence. Parents are not parenting in isolation, but in response to several factors including child traits [60]. Thus, the parent-child relationship is bi-directional and additional studies are needed to understand causal pathways between parenting behaviors and child physical activity that include measures of specific child behaviors (e.g. temperament) or potential mediating variables (e.g. self-efficacy).

Overall, this work suggests that moderating factors such as parenting style should be considered in future studies of parental influences over child physical activity levels. In general, a more focused investigation of theoretical or methodological moderators in addition to mediators and confounders may clarify previously indeterminate findings across investigations [7, 61]. For instance, parental reinforcement of child physical activity may be moderated by parenting style, but may be mediated by children's efficacy and competency [7]. It would also be interesting to include measures of child report of parental behavior to determine the similarities and differences with parental report of their own behavior. How children perceive their parents - as supportive or unsupportive - may be an important factor and whether differences exist for parent-child gender pairs. Other factors such as family cohesion, parental engagement, and parent-child communication have been evaluated in a few studies but more work is needed [27]. Additionally there may be other domains of parental support that have not yet been evaluated in the literature (e.g. the extent to which parents alter the home environment to be supportive of child physical activity) [62] and whether parental influence may have an impact on weekday versus weekend activity levels for children.

Currently, interventions have demonstrated mixed results and the effectiveness of family involvement methods or parental components in promoting physical activity in children remain unclear [63]. One reason may be that interventions have not accounted for certain factors - such as parenting style - which may alter the effectiveness of the parenting practices targeted in the intervention [64]. Development of theory-specified, empirically verified models that predict family and parent influences on children's physical activity behaviors are needed to develop effective interventions.

Overall, this work extends the current literature by demonstrating the potential moderating role of parenting style on the relationship between activity-related parenting practices and children's objectively measured physical activity while controlling for known covariates. Our findings add to the growing evidence base, which suggests that parents need to be aware of their behaviors that may or may not have unintended consequences on their child's health [65]. Future studies in this area are warranted to identify the mechanism by which parent's influence their child's physical activity behavior.

This research was supported by awards from the Centers for Disease Control and Prevention (Public Health Dissertation Research Award, Grant 1R36DP001325-01) and New Balance Foundation with additional funding provided by Save the Children, US Programs and the Robert Wood Johnson Foundation (Grant 59458).