As with any study, limitations must be acknowledged. This study was cross-sectional; as such we cannot know the casual pathways between the predictor variables and behavior change. The study design allowed for internal reliability analysis but does not allow for analysis of data validity. Parents reported behavior change over the past year and
potentially may not recall changes that were attempted but not sustained during this time period. Social desirability may have resulted in parents over reporting behavior change as well as not accurately reporting their height and weight. Finally, study findings may not be generalizable to other populations. Specifically, participation in free/reduced lunch was a proxy measure for family SES and this variable was the only significantly different characteristic between families who returned complete surveys and families who returned uncompleted surveys. Study participants were also recruited from an Appalachian, predominately rural setting. As a result, additional research
is necessary to determine whether the findings are appropriate for urban settings or in other areas of the country. Mothers were the primary respondents; additional recruitment targeting fathers would be necessary to more fully understand how parental perceptions and concerns might influence behavior. Despite these limitations, this study adds to the
literature by including multiple parent weight related variables as predictors of child behavior change and multiple child weight-related variables as predictors of parent behavior change. Whereas previous studies have generally focused on children
who were overweight or obese,22,38 this study included children whose BMI percentile was less