- Top of page
What predicts whether a child will be at risk for obesity? Prior studies have often focused on the food habits and eating behaviors of children [1-4]. However, a wide range of family-related factors also influence how a child eats [5, 6], including social facilitation and consumption norms . While social facilitation has often concentrated on mimicry in how much one eats  or what one prefers , we expand it here to examine a wider range of behaviors and rituals that could influence the social dynamics during dinner time. There has been recent emphasis on obtaining a better understanding of behavioral determinants of obesity . Finding what repeated behaviors or family rituals might unknowingly correlate with child or adult body mass index (BMI) could suggest promising dinnertime interventions to address childhood obesity .
The long-ranging impact of dinners on a child's health and nutrition intake can be either positive, as in the case of higher diet quality and healthful meal patterns , or negative, in the case of bulimia nervosa . In many cases, the impact of these childhood dinner experiences on long-term behavior may influence feelings of psychosocial well-being as much as nutrition . Even what one simply talks about at mealtime has been shown to support literacy development .
Three general categories of dinner-related behaviors or rituals are suggested by the literature: (i) preparation involvement before dinner, (ii) the social dynamics during dinner, and (iii) the dining environment itself. Preparation involvement during dinner—cooking with kids—has received recent attention as a way to improve a child's understanding, appreciation, and talent toward meal preparation, as well as a way to encourage family bonding. A large part of the hopeful success for preparation involvement with children is the success this has received in classroom situations [15, 16]. Unfortunately, there is little research to guide us as to what impact cooking with kids would have on a child's BMI or on long-term development.
When examining the social dynamics of the dining environment, there is a wide range of research that suggests that family meals correlate with psychosocial well-being among adolescents  and that the type of things a family discusses may also play a role in one's development . The link between the social dynamics of family conversations and BMI has not yet been made, but it seems that if there is a link between dinner conversation and positive psychological development , and if there is a link between positive psychological development and positive eating habits , there may be a relationship worth exploring. In particular, how might positive social dynamics—such as positive topics of dinner conversation—relate to the BMI of a child?
Last, the dining environment itself has recently been identified as a potentially overlooked influencer of food intake and weight gain. Many of these studies have focused on the tablescape : how the size and color of plates , serving bowls , and spoons  influence how much both adults and children consume. What has been hypothesized [22, 23] but less carefully investigated in a family situation is where they eat (kitchen or dining room vs. in another room) and if watching television during dinner has any impact on their BMI [24, 25].
There is a rich sociological history surrounding the ritual of family dinners . Based on Fiese and colleagues , we define dinner rituals as the set of patterned interactions and routines surrounding meals that are repeated over time. It would be useful to know which of these rituals are positively or negatively correlated with a child's BMI. Knowing this information could provide useful intervention points that could be either investigated in more detail or could be used as easy, exploratory intervention points. That is the purpose of this research.
- Top of page
For the purpose of the study, we analyzed unreported data from an unrelated study on snacking behavior . Participants were 201 (115 females) children entering the third to sixth grades in the fall of 2011 from the Chicago Metro area. Data reported in this article was available for 190 parent participants (173 females) and 148 children participants (87 girls). A $70 incentive was provided to the parents of participating children. The study had Cornell University's Institutional Review Board approval, and consent from both the parents and children was obtained prior to the study.
Because it was feared that most parents do not have an accurate assessment of their child's weight and height, it was necessary for them to bring their child to the research facility where their height and weight could both be accurately measured. This was done following the completion of the survey to reduce the chance of any confounding with answers of their parent. At that point, the height and weight of the parent was also taken.
Parents and their children were recruited and met in small groups from 5 to 11 participants. After arriving at the facility and being told that they would be completing a survey on family dining and eating behavior, the children were taken to an adjoining room where they could watch television and eat a snack while their parents completed the survey.
The questionnaire completed by parents concerned the whole family's mealtime habits. They were asked a broad series of questions concerning how many days they engage in mealtime activities, such as discussing their day, during a typical week. We focused on questions regarding both the food environment and the eating environment , along with social interactions, meal involvement, the dining environment, and the dietary habits of both the parents and the children. Most questions were either frequency-based (e.g., “How many times in the past week did you or your family eat dinner in the kitchen or at the dining room table?”) or asked as Likert questions (using a nine-point scale to show the extent to which they strongly disagreed  or strongly agreed  with a particular statement, such as “My child usually helps me clean up after dinner.”). Specifically, because family involvement is often characterized as the interactions between family members, parents were also asked how likely family members are to “talk meaningfully about their day,” “compliment each other at dinner,” and “compliment the person who made the dinner.” Demographic information included parents' age, gender, and educational level (ranging from 1—some high school to 5—graduate degree).
Pearson correlations were calculated to assess the correspondence between the dining variables and BMI. Next, a multiple regression analysis was carried out to simultaneously test the effect of social dynamics, meal involvement, and dining environment variables on BMI. The educational level of parents was included in the regression models to control for its influence. When BMI was calculated, we excluded one female adult with a BMI of 12, which was most likely a data-entering mistake.
- Top of page
The average age of the sample of parents was 41.0 years (SD = 7.0) and 102 of the 186 parents (55%) who responded to the educational level question had at least a college or graduate degree. As can be seen in Table 1, the BMI of all adults was related to dining rituals in similar ways. Talking meaningfully about children's day correlated negatively with the BMI of all adults (r = −0.18, P < 0.05) and BMI of mothers (r = −0.20, P < 0.01). The highest negative correlation was between eating in the kitchen or in the dining room and parents' BMI (parents' BMI, r = −0.31, P < 0.01; mothers' BMI, r = −0.32), P < 0.01). In addition, BMI was negatively correlated to remaining at the table until everyone is finished eating (parents' BMI, r = −0.17, P < 0.05; mothers' BMI, r = −0.19, P < 0.05). By contrast, eating with the TV on was positively correlated with BMI (parents' BMI, r = 0.23, P < 0.01; mothers' BMI, r = 0.22, P < 0.01). In other words, parents with a higher BMI indicated more frequent eating with the TV on.
Table 1. Dinner ritual correlates of BMI
| ||Parent's BMI (n = 190)||Mother's BMI (n = 173)||Child's BMI (n = 148)||Girl's BMI (n = 87)||Boy's BMI (n = 61)|
|Social dynamics|| || || || || |
|Saying grace or prayer with dinner||0.09||0.09||0.05||0.05||0.04|
|Talk meaningfully about their day||−0.18*||−0.20**||−0.13||0.01||−0.24|
|Argue at dinner||0.02||0.05||−0.01||0.06||−0.05|
|Compliment each other at dinner||−0.03||−0.02||−0.02||0.05||−0.07|
|Compliment the person who made the dinner||−0.06||−0.06||−0.09||−0.02||−0.21|
|Meal involvement|| || || || || |
|Cook dinner with your child||0.09||0.11||0.13||0.23*||0.09|
|My child is usually involved in helping make dinner||0.03||0.04||0.11||0.26*||0.04|
|My child usually helps me clean up dinner||−0.14||−0.14||−0.01||0.11||−0.08|
|Dining environment|| || || || || |
|Eat in the kitchen or dining room||−0.31**||−0.32**||−0.24**||−0.35**||−0.11|
|Eat with the whole family||−0.10||−0.09||−0.05||0.06||−0.14|
|Eat with the TV on||0.23**||0.22**||0.03||0.07||−0.02|
|Remain at the table until everyone is through eating||−0.17*||−0.19*||−0.16*||−0.001||−0.31*|
|Prepare different foods for self versus kids||−0.05||−-0.03||−0.04||−0.10||0.01|
Moderate correlations were observed for children's BMI in regards to eating in the kitchen or dining room (r = −0.24, P < 0.01) and remaining at the table until everyone is finished with their meal (r = −0.16, P < 0.05). Interestingly, the BMI of girls was associated with meal involvement (“cook dinner with your child,” r = 0.23, P < 0.05; “my child is usually involved in helping make dinner,” r = 0.26, P < 0.05), but not with social dynamics (all rs = ns). Eating in the kitchen or dining room was most strongly associated with girls' BMI (r = −0.35, P < 0.01). For boys, a negative association was observed between remaining at the table until everyone has finished eating and their BMI (r = −0.31, P < 0.05).
Results of the regression analysis using the same set of independent variables are shown in Table 2. The proportion of variance of the dependent variable BMI explained by the independent variables was 20% regarding all adults' BMI and 23% regarding mothers' BMI. After controlling for the education level of parents, eating in the kitchen or dining room was the strongest independent predictor of BMI across models, except for boys. This variable was the only significant predictor of BMI of all children (β = −0.27, P < 0.01) and girls (β = −0.42, P < 0.01), and for parents (β = −0.31, P < 0.01) and mothers (β = −0.34, P < 0.01) as well; the more frequently they report eating in the kitchen or dining room, the lower their BMI.
Table 2. Dinner ritual and education regression coefficients related to BMI
| ||Parent's BMI (n = 190)||Mother's BMI (n = 173)||Child's BMI (n = 148)||Girl's BMI (n = 87)||Boy's BMI (n = 61)|
|Social dynamics|| || || || || || || || || || |
|Saying grace or prayer with dinner||0.15||2.07*||0.17||2.27*||0.08||0.90||0.11||1.00||0.01||0.06|
|Talk meaningfully about their day||−0.14||−1.57||−0.14||−1.54||−0.09||−0.81||0.11||0.75||−0.29||−1.42|
|Argue at dinner||0.09||1.33||0.15||2.03*||0.07||0.85||0.10||0.88||0.04||0.30|
|Compliment each other at dinner||0.11||1.26||0.11||1.22||0.05||0.50||0.07||0.54||0.15||0.77|
|Compliment the person who made the dinner||−0.10||−1.15||−0.10||−1.11||−0.12||−1.16||−0.13||−0.94||−0.26||−1.57|
|Meal involvement|| || || || || || || || || || |
|Cook dinner with your child||0.19||2.08*||0.24||2.57*||0.13||1.17||0.12||0.77||0.14||0.78|
|My child is usually involved in helping make dinner||−0.14||−1.50||−0.18||−1.86||0.01||0.09||0.13||0.80||0.04||0.23|
|My child usually helps me clean up dinner||−0.01||−0.11||−0.05||−0.58||0.04||0.35||0.14||1.06||−0.03||−0.15|
|Dining environment|| || || || || || || || || || |
|Eat in the kitchen or dining room||−0.31||−3.77**||−0.34||−3.94**||−0.27||−2.74**||−0.42||−3.42**||−0.10||−0.47|
|Eat with the whole family||0.09||0.98||0.13||1.31||0.05||0.49||0.004||0.03||0.06||0.29|
|Eat with the TV on||0.07||0.88||0.09||1.12||−0.11||−1.16||−0.16||−1.11||−0.12||−0.68|
|Remain at the table until everyone is through eating||−0.06||−0.71||−0.05||−0.53||−0.11||−1.09||−0.12||−0.87||−0.09||−0.40|
|Prepare different foods for self versus kids||−0.10||−1.39||−0.12||−1.51||0.002||0.03||0.03||0.25||−0.09||−0.46|
|Control variable|| || || || || || || || || || |
|Educational level of parents||−0.24||−3.17**||−0.22||−2.93**||−0.14||−1.52||−0.01||−0.10||−0.19||−1.13|
|Adjusted R2|| ||0.20|| ||0.23|| ||0.05|| ||0.08|| ||−0.04|
The more frequently parents (β = 0.19, P < 0.05) and mothers (β = 0.24, P < 0.05) reported cooking with their child, the higher their BMI. For these groups, saying grace or prayer with dinner positively influenced BMI (parents β = 0.15, P < 0.05 and mothers β = 0.17, P < 0.05). For girls' BMI, the value of the adjusted R2 was 0.08, showing that these variables accounted for 8% of the variation in BMI score. Boys' BMI was not well explained by the data as all variables were statistically marginally significant or insignificant with a negative adjusted R2.
- Top of page
By focusing on family dining rituals, this research departs from the more food-centric approaches --such as parental feeding practices and family meal patterns-- that have been used to predict BMI. These dinnertime rituals often tended to correlate with either parental BMI or the child's. This starts with meal preparation. While being involved in meal preparation was unrelated to the BMI of young boys, it was positively correlated with the BMI of young girls. There is no effort to imply causality. We do not know if cooking leads girls to become heavier or whether heavier girls are more interested in learning to cook and helping with meal preparation.
An interesting finding with young boys was that the social dinner dynamics of the family was correlated with lower BMI. In families where children remained at the table until everyone was finished eating, the BMI of boys was lower than those who did not remain at the table. Another social finding was that in families who claimed to talk meaningfully about their day, parents had significantly lower BMIs than those families which did not. This general finding is important because it may suggest the potential importance of social dynamics that can occur at dinner. At the very least, it underscores the potential danger of relying on television instead of meaningful conversation because eating with the TV on was significantly related to the BMI of parents. In fact, eating anywhere other than the kitchen or dining room was related to higher BMIs in both parents and in children. The regression analysis similarly showed that this behavior is a strong predictor of BMI of both parents and children.
To summarize, the ritual of where one eats and how long one eats seems to be the largest driver of both adult and child BMI. Eating in the kitchen or at the dining room table and remaining at the table until everyone is finished eating were both associated with lower BMIs for parents children. Such behavior may be related to less distracted eating or less supervised eating. These behaviors might also be associated with stronger or more positive feelings with others that might supplant a tendency toward overeating.
Limitations and future research
There is a rich sociological history surrounding the ritual of family dinners, and some of these rituals appear to correlate positively or negatively with a child's BMI. This study preliminarily investigates possible relationships that relate to this dining ritual. We corrected for educational level of the parents but did not include other relevant (sociodemographic) variables, such as income and ethnicity. It will be important for future studies to take account of these variables and identify underlying mechanisms explaining results. Although our results are correlated, they provide a strong justification for further research in this area.
There is a promising opportunity to extend a study like this to investigate how these dining rituals influence children of different ages—toddlers to teens. Additionally, another opportunity would be to survey adults about their memory of their childhood dinner rituals and correlate these with their current BMI. Although such a retrospective approach would not be without limitations, in combination with these findings, it could help narrow the potential points of influence for increased BMI.
Implications and conclusion
Although the results are correlated, they suggest the overwhelming importance of the social interaction of a meal—eating at a table and having a positive interaction. Dinnertime rituals are not costly to change, and this research suggests that families may wish to consider experimenting with different rituals. Eating together at a kitchen table or dining room with the television off appears to be the best place to start. Additionally, encouraging the children to talk meaningfully about their day might also be an easy change to make. One approach is to have everyone at the table share four items about their day: (i) their high point, (ii) their low point, (iii) who they appreciate most and why, and (iv) their plan for tomorrow.
The dinner table may be more than simply a place where food is eaten. It may be where social support and family involvement come together—both of which may relate to the BMI of children, as well as parents. Change at the dinner table might be a simple intervention to fight future family fat.