Environmental and cultural correlates of physical activity parenting practices among Latino parents with preschool-aged children: Niños Activos

Latino children carry a disproportionate burden of the obesity epidemic, starting at a young age [1]. Physical activity (PA) was negatively associated with adiposity in children [2, 3], supporting that PA plays an important role in pediatric obesity prevention. It is therefore imperative to identify modifiable factors that influence Latino preschool children’s PA. Parents are an important social influence that affect Latino preschoolers’ PA [4]. Parenting practices (context specific parent behaviors intended to influence their child’s behavior) [5] for PA were linked to children’s PA [6]. PA parenting practices to support children’s PA were positively associated with school aged children’s PA [7, 8]. For preschool aged children, parental use of PA to reward and control behavior, and logistic support for PA were associated with objectively measured child MVPA [9]. In a sample of Latino parents of preschoolers (n = 85), psychological control and registering children for sports was significantly positively associated; while promoting screen time was negatively associated with children’s objectively measured moderate PA [10]. Latino parents of preschoolers reported using parenting practices that both encouraged and discouraged PA [11].

From a Social Ecological perspective [12‐14], multiple social and environmental factors should influence behaviors such as PA parenting practices. A few studies have examined social ecological correlates for supportive PA parenting practices [15, 16], but not for parenting practices that may discourage PA among preschoolers, nor specifically among Latino parents. Research found that characteristics of the neighborhood were associated with children’s weight status [3] and PA [17‐23], but it has not been established whether neighborhood characteristics are also associated with PA parenting practices. This would provide a potential mechanism through which the neighborhood characteristics may influence children’s PA, especially among younger children when parents are most influential.

Culture may also influence PA parenting practices. Mexican American mothers had different beliefs about parenting than European American mothers during pregnancy, which was influenced by their level of acculturation [24]. Acculturation also influenced Latino parenting in other contexts [25, 26], but it is not known if cultural factors influence Latino parenting practices specific to children’s PA.

Identifying correlates of PA parenting practices that fit within the Social Ecological framework [27, 28] can help detect targets for future interventions and policy changes to promote greater use of parenting practices that encourage PA and to decrease the use of those that discourage PA. The aim of this study was to examine the contributions of socio-demographic, cultural, objectively-measured and parent perceived environmental factors to explain variation in parenting practices that encourage or discourage PA among Latino preschoolers.

A cross-sectional study of Latino parents (n = 240), Niños Activos (“Active Children”), was conducted in Harris County (Houston), TX with data collected from July 2011 to March 2012. Data were analyzed in 2013. Details about the study have previously been reported [10], and are briefly described here. To ensure variability of the neighborhood environmental variables, parent recruitment was stratified by objective traffic and crime risk characteristics of families’ neighborhoods at the census block level [10]. Traffic risk indices were calculated based on traffic related injuries and motor vehicle miles traveled for each census block group in Harris County from 2004–2008 [10]. Crime index data, based on FBI Uniform Crime Report data from 1998–2007 at the census block group level for Harris County, were included (Tetrad Inc, Vancouver, British Columbia; http://​www.​tetrad.​com) [10]. Using median splits of the crime and traffic safety index scores, each block group within Harris County was classified as high crime/high traffic; high crime/low traffic; low crime/high traffic; or low crime/low traffic. The goal was to enroll about 60 participants from each type of block group. Recruitment from census block groups with high crime and low traffic (typically in the outskirts of Harris County) proved difficult due to lack of infrastructure. After we maximized enrollment from the high crime and low traffic block groups and 60 participants were recruited for the three other types of block groups, we employed convenience sampling to achieve the goal of 240 total participants. Figure 1 depicts final recruitment stratification.

In general, parents’ perception of their neighborhood’s physical and social environment had the greatest effect on the variance of the different types of PA parenting practices, while objectively measured characteristics of the neighborhood had little to no impact on the PA parenting practices measured. Specifically, parental perceptions of perceived physical and social disorder, traffic safety, availability of places for child’s PA in the neighborhood, and presence of neighborhood social informal control were positively associated with parental engagement/structure for promoting child PA. Parents who perceive their neighborhood to have low traffic risks, plenty of places for children to engage in PA, and adequate neighborhood informal social control may feel more comfortable engaging with their child in PA. Alternatively, parents who believe it is important to engage in PA with their child, may self-select to reside in neighborhoods with such characteristics. It is not self-evident why parents who perceive more physical and social disorder in their neighborhood provided greater engagement and structure to promote PA. It is possible that parents who perceive neighborhood social disorder may feel they need to take a more active, engaging role with their child in order to provide supervision for their child.

Places for child PA was also positively associated with registering the child for sports or dance, possibly due to these locations having more opportunities for registering children in PA activities (e.g. sports or dance). Having outdoor toys available was positively associated with availability of active play equipment (as expected) and neighborhood informal social control. Parents who perceived greater traffic hazard in their neighborhood and stranger danger also reported more use of discouraging PA for safety concerns, as would be expected. The availability of active play equipment was negatively associated with safety concerns, likely due to the fact that parents make such toys available only if they believe outdoor play to be safe. Lastly, parental report of stranger danger was positively, while places for child’s PA was negatively associated with psychological control respectively. This is one of the first studies that has investigated psychological control as a practice to impact children’s PA and warrants further study to explore these findings.

In addition to parental perceptions of their neighborhood influencing their use of PA parenting practices, a few sociodemographic variables were also significant correlates. For example, parents with older children were less likely to promote inactive transport. This sub-scale included items such as carrying the child, and pushing them in a stroller instead of letting them walk, which parents are likely to do less as the child gets older. Parents who lived in an apartment were less likely to have outdoor toys available, which may be due to space limitations, but more likely to use psychological control.

Only a few other studies have investigated correlates of PA parenting practices [15, 16]. One investigated child and parental characteristics as correlates of PA parenting to restrict sedentary time, monitor, and stimulate PA among 5 year old children in the KOALA birth cohort [16]. Unlike that study from Europe, we did not find that child gender was significantly associated with PA parenting practices, nor that lower education among parents was associated with less use of practices to stimulate PA. Instead, in this Latino sample, higher parental education was associated with less frequent use of Engagement/Structure for PA. The second study [15] focused on family and community factors. Their combined model explained 37% of the variance of supportive PA parenting practices among a sample of low-income, mostly non-Hispanic white or black parents of preschool-aged children. Unlike this study, community factors were not significantly associated with PA parenting practices in their multivariate models. This may be due to different instruments being used to assess PA parenting; different neighborhood characteristics assessed, and populations of different ethnicities being sampled.

Characteristics of the neighborhood where children reside have been associated with children’s PA [17‐21, 23], but not consistently [40, 41]. The current study provides evidence that Latino parents’ perception of their neighborhood’s environment influenced how likely they were to encourage or discourage PA among their preschool-aged child, which may in part mediate the correlation of neighborhood characteristics with children’s PA. In this sample, the perceptions of their neighborhood were more influential than objectively measured neighborhood characteristics. This is contrary to a review that found objectively measured neighborhood attributes to be more consistently associated with youth’s PA than perceived neighborhood characteristics [22]. It is possible that parental perceptions of the neighborhood are more important for influencing parental behaviors to encourage or discourage PA among younger children, than older children’s PA behaviors. Youths’ PA may be more directly influenced by the objectively measured attributes, such as distance to a park.

Acculturation was correlated with PA among Latino adults, with greater acculturation associated with more PA in most Latino adult studies [42]. The associations of acculturation and PA among youth is less clear [43‐47]. In the present study, cultural variables only had a weak main effect on PA parenting practices, specifically on discouraging PA due to safety concerns.

This study had several strengths including a sample at high risk for obesity who may be less active; stratified sampling to get a varied neighborhood environment exposure; assessing PA parenting practices that both encourage and discourage child PA with a validated instrument [10]; and using previously developed validated scales to assess neighborhood characteristics and cultural variables whenever possible. There were also limitations to consider. The study only sampled from one US city, and may not be representative of other Latino communities. Due to response burden we did not assess psychological variables, such as parental stress and depression, previously associated with PA parenting [15]. Only three objective neighborhood characteristics were assessed, and for the traffic and crime variables, the data were a few years old. This is a problem when relying on data that are only released intermittently and may not accurately reflect the current status of neighborhoods that have recently undergone a revitalization or decline.

Primarily socio-demographic and perceived environmental variables significantly contributed to explaining 6.8–38.9% of the variance of PA parenting practices. Parental perception of their neighborhood appeared to be the most significant correlates. These findings suggest that interventions to promote PA among young children and target parents as agents of change need to consider multiple factors to effectively change parent’s behaviors to encourage and not discourage PA among their preschool aged children. This may be achieved by tailoring or adapting interventions to specific family social-ecological contexts.