Home food and activity assessment. Development and validation of an instrument for diverse families of young children

Highlights

• We tested a home food and activity instrument with diverse families

• Homes often contain unhealthy snack foods and low amounts of fruits and vegetables

• Half of the homes reported a television in the preschool child's bedroom

• Low education and income families report less healthy home characteristics

Abstract

The purpose of this study was to refine and psychometrically test an instrument measuring the home food and activity environment of geographically and economically diverse families of preschool aged children. Caregivers of preschool aged children (n = 83) completed a modified self-report questionnaire. Reliably trained researchers conducted independent observations on 25 randomly selected homes. Agreement statistics were conducted at the item level (154 total items) to determine reliability. Frequency counts were calculated to identify item availability. Results showed Kappa statistics were high (.67–1.00) between independent researchers but varied between researchers and parents resulting in 85 items achieving criterion validity (Kappa >.60). Analyses of reliable items revealed the presence in the home of a high frequency of unhealthy snack foods, high fat milk and low frequency of availability of fruits/vegetables and low fat milk. Fifty-two percent of the homes were arranged with a television in the preschool child's bedroom. Physical Activity devices also were found to have high frequency availability. Families reporting lower education reported higher levels of sugar sweetened beverages and less low-fat dairy (p < .05) compared with higher education families. Low-income families (<$27K per year) reported significantly fewer Physical Activity devices (p < .001) compared with higher income families. Hispanic families reported significantly higher numbers of Sedentary Devices (p < .05) compared with non-Hispanic families. There were no significant differences between demographic comparisons on available fruits/vegetables, meats, whole grains, and regular fat dairy. A modified home food and activity instrument was found to reliably identify foods and activity devices with geographically and economically diverse families.

Introduction

The prevalence of obesity has reportedly plateaued for preschool aged children (2–5 years) according to recent epidemiological estimates (Ogden, Carroll, Kit, & Flegal, 2012). Specific socio-demographic groups of children at all ages remain consistently higher in body mass index (BMI) compared with others. For instance, preschoolers from families who identified as being Mexican American or non-Hispanic black were higher in prevalence for overweight (> 85th to 94.9th BMI percentile) and obesity (>95th BMI percentile) compared with non-Hispanic white preschoolers (Ogden et al., 2012). In addition to race and ethnicity, low socioeconomic status (SES) has been related to an increased risk for obesity during childhood (Shrewsbury & Wardle, 2008). Socioeconomic status may impact lifestyle behaviors and environmental factors, including food access and types of physical activity (Wang & Lim, 2012). While the causal relationship between SES and childhood obesity is considered complex, children from disadvantaged backgrounds generally have greater levels of obesity than children from advantaged backgrounds (Sobal, Stunkard, 1989, Wang, Lim, 2012).

Diversity in geographical location (rural versus urban) represents an additional demographic focal point for obesity risk. Approximately 30 years ago, the prevalence of childhood obesity appeared higher in large, urban areas compared with rural areas (Dietz & Gortmaker, 1984). However, more recent data show an opposite finding in which rural children (and adults) are now more likely to be overweight or obese (Liu et al., 2012). The mechanisms of rural and urban differences in obesity prevalence are poorly understood and possibly moderated by SES factors. In particular, rural residents are more likely to be older, less educated, and have a lower income; factors known to associate with higher rates of obesity (Lantz et al, 1998, Martikainen, Marmot, 1999, Miller et al, 1987). Structural challenges that may impact weight in rural communities include fewer services for nutrition education (e.g., access to dietitians) and physical activity (e.g., exercise facilities) as well as treatment facilities for weight management (Tai-Seale & Chandler, 2003). Thus, living in rural communities has become a health disparity (Gamm, Hutchison, Bellamy, & Dabney, 2002).

Little evidence is available, however, on validated tools identifying characteristics of the home food and activity environment among families with SES and geographical risk factors for obesity (Boles et al, 2013, Ostbye et al, 2013, Pinard et al, 2012). This limitation in etiological understanding can be characterized within a social–ecological theoretical view, in which obesity may result from an interactive, bidirectional influence of multiple weight-affecting factors at the individual level (e.g., temperament), within the family (e.g., feeding practices) and home environment (e.g., available foods and sedentary devices), to the community, including schools and neighborhoods (e.g., access to parks or corner stores); (see Story, Kaphingst, Robinson-O'Brien, & Glanz, 2008 for review). While studies on the home physical environment of young children have emerged in the last decade, most have focused on white families and underrepresented families from low education and income backgrounds, the families most at risk for obesity development (Boles et al, 2013, Bryant et al, 2008, Fulkerson et al, 2008, Ostbye et al, 2013, Spurrier et al, 2008). Previously published home assessments may have been conceptually limited by the types of foods and activities included in them and may not have represented foods and activities that are available for families with socioeconomic and racial and ethnic diversity. As a result, important considerations related to methodological and psychometric testing remain largely unexplored with families of SES and geographical diversity.

Pinard et al. (2012) conducted a systematic review of forty home environment measures. They noted a lack of reported instrument psychometric properties and a narrow focus (food or physical activity; not combined). Additionally, few measures were inclusive of families from geographic and socioeconomic diversity, those most at-risk for development of obesity and thus, generalizability may be limited. The present study aimed to address these limitations by developing and testing a comprehensive home environment tool for families of diverse SES and geographical characteristics, known factors to be associated with increased risk for obesity.