weight control as part of a balanced diet, as FV’s increase feelings of satiety and may reduce intake of high-energy foods [9,18]. However, observational data evaluating the relationship between FV intake and BMI provide inconclusive results [9,18,19]. A systematic review of the literature suggests
Table 3. Differences in BMI based on fruit/vegetable availability in mothers responding to an annual survey regarding fruits/vegetable in health (n = 1463).
Table 4. Mother’s fruit/vegetable intake versus mother’s report of child’s body shape in mothers responding to an annual survey regarding the role of fruits/vegetables in health.
Figure 2. Comparison of BMI of mother’s who selected images that most closely resembled their child’s body shape. BMI values represent mean ± SD. *Significantly different from the BMI of mothers who selected images A/B/C. P trend = 0.001.
that there may be a relationship between low FV intake and higher BMI ; however, most studies that assess this relationship are designed with endpoints of FV intake and chronic disease outcomes , making it difficult to accurately extrapolate their findings to obesity, specifically. Our study did show a statistically significant relationship between fruit availability and BMI in mothers. These findings are consistent with much of the research regarding home FV availability and weight status, most of which has focused on children and adolescents and not adults [21,22]. Interestingly, this effect was not seen with vegetable availability and BMI, as mothers who selected that vegetables were “occasionally/rarely/ never available” in their home had lower BMIs than those who selected that vegetables were “almost always/ usually available.” Supporting our data, an analysis from the USDA’s 1994- 1996 Continuing Survey of Food Intakes by Individuals (CSFII) found that lower BMI was associated with greater fruit, but not vegetable, intake in adults and children . Although we did not see a significant difference in BMI between women selecting the highest and lowest vegetable availability categories, it should be noted that women who indicated that vegetables were always available in their home had a lower BMI than those who did not choose this option. In addition, the BMI of this group was similar to the BMI of the group who indicated that fruits were always available in the home. One important issue that may potentially explain these results is the very low number of women who selected the lowest categories of vegetable availability (n = 74) compared with those who indicated that vegetables were always available (n = 753). Furthermore, we did not obtain information regarding fruit and vegetable preparation methods. Vegetables consumed may have been fried or eaten with high calorie sauces or butter. This effect may explain the lack of an association between vegetable availability and BMI.
According to the social ecological theory, there are many factors that influence children’s weight status, including parental behaviors and beliefs [12,15]. Our study explored the relationship between mother’s BMI and child’s (aged 2 - 12 years) body shape. Child’s body shape was used as a proxy for child’s BMI, as validated by Saxton et al. . We found a significant association between mother’s BMI and child’s body shape. Women who selected that their child resembled an image of a larger body size had higher BMIs than those who selected a smaller body size. These findings agree with previous reports regarding parental BMI and child’s weight status [25,26].
In addition to the relationship between mother’s BMI and child’s body shape, we explored additional variables that may influence child’s weight status. Previous work suggests that FV availability in the home is related to children’s FV intake . Children’s FV intake may then confer benefits such as healthier weight status and reduced adiposity [18,22]. Despite this rationale, we failed to observe a relationship between FV availability and child’s body shape. However, actual intake of FV in children was not assessed.
There is a large body of evidence exploring the relationship between mother-child FV intake resemblance [13,14,27] as well as FV intake and BMI in both adults and children [18,19]. To our knowledge, this is the first study examining the relationship between maternal FV intake and child’s body shape. We observed a significant inverse association between mother’s fruit intake and child’s body shape, suggesting that lower maternal fruit intake may be associated with her child’s larger body size. This relationship did not exist between vegetable intake and child’s shape, perhaps for the same reasons that it did not exist between mother’s BMI and home vegetable availability. These findings provide further evidence suggesting that interventions to improve FV intake in children should also focus on promoting FV intake in mothers, as the family environment is a strong determinant of behavior development .
This is the first internet-based study to target a specific generation of mothers and for which we have novel information regarding mother’s perception of child’s body size to evaluate the role of FV in family health. The study was large with responses from over 1400 mothers. However, the study does have limitations. This study was conducted using cross-sectional data, limiting the ability to determine causation. Additionally, this study was conducted online and is therefore biased towards women with internet access. This study was also based on selfreported height and weight, FV intake, and FV availability; thus, respondents may have provided inaccurate information. Without a marker of validation (height and weight measured by a trained professional, a biomarker of intake such as serum carotenoids, or in-home appraisal of FV availability), we cannot ascertain the validity of self-reported data in the current study. We also did not have any information regarding other aspects of the diet, making it difficult to control for other potential confounders. Furthermore, social desirability may have resulted in answers that were not truly representative of actual behaviors and beliefs.
Additionally, our survey collected information on mother’s perceptions of her child’s body shape. Although these images were designed using BMI percentiles developed over 30 years ago, the scale itself was not developed until 2002  and currently represents the most up-to-date scale of children’s body images. Furthermore, this scale is unique as each image is an actual photograph of a child that correlates to a specific BMI, whereas previously developed scales simply depict silhouettes that do not necessarily correspond to actual BMI values . A validation study by Saxton et al. in 2009 indicated that children were able to identify their body shape using these images with reasonably high accuracy . However, there is literature that indicates that mothers tend to underestimate their child’s body size [28,29]. Currently, 16.9% of American children are classified as obese , while only 5.9% of women in this survey chose the body images that most closely correlate with obesity (F = 90th percentile; G = 97th percentile). This suggests that some women may have misclassified her child or that the sample of women surveyed is not largely representative of the American population.
The relationships between mother’s BMI, FV intake, FV availability, and child’s weight status are extremely complex and not well understood. Using a national sampling of mothers, we were able to evaluate FV availability in the home in relation to the BMI of the mother and also the body size of the child. The current study supports a role of fruit and, to a lesser degree, vegetable availability in BMI of mothers. Our data suggest a relationship between mother’s fruit, but not vegetable, intake and child’s body shape. The evidence generated from this study, particularly that regarding women’s BMI and home FV availability, and mother’s FV intake and child’s body shape, should be further explored in longitudinal studies in order to inform on future interventions targeting FV consumption and healthy weight in the family setting.
The authors would like to acknowledge OnResearch Inc. for their efforts in survey design as well as the United States Department of Agriculture’s National Need Fellowship that provided the necessary training for completion of this research.
*The authors have no conflicts of interest to report.
Funding for the survey was provided by the Produce for Better Health Foundation.