“Do as I do”
Modeling behavior to effect dietary change in children

Jacob Schor ND, FABNO
November 16, 2013

An interesting paper was published last September that describes an experiment to improve the eating habits of children. You might be able to guess the findings but seeing them in print has given me pause.

The study was randomized controlled trial for designed to evaluate the effectiveness of several school-based interventions for increasing children’s fruit consumption.
Pretty much everybody knows by now that eating more vegetables and fruits is better for your health. The question though is how to get people and in particular children to do this.

The dietary intakes of the kids was measured through a simple 2-day dietary diary before the intervention began in October 2008 and then at the end of the intervention period (June 2009) and at 1-year follow-up (June 2010).

In all, two hundred and eighteen 9-year-old elementary students were randomly assigned to one of three groups. Children in two groups were exposed to the experimental interventions aimed at increasing fruit consumption and the third group served as a control.

There were two interventions tested.
The first was an educational curriculum designed to provide general information about a healthy lifestyle and to build skill for choosing healthful foods. The curriculum consisted of 29 lessons of 15 minutes each delivered to the class by their own teacher once a week for the entire school year. The teachers were trained and provided with a curriculum and teaching materials.

The second experimental group received no educational training, instead they were exposed to their teacher eating healthy snacks during the school day, in particular fruit. The teachers merely acted as role models, bringing fruit to school at least three times a week. They kept the fruit on their desks and ate it during a break in the day. The control group of students received no intervention, neither educational or exposure.

Outcome Measures: Parents of the children completed 2-day dietary records listing the foods consumed at home or which the students recalled eating away from home, during two consecutive weekdays before the start of the intervention, after the intervention was complete and again one year later. Researchers measured the number of servings of the selected food groups in the food records with particular attention to fruit consumption.

Fruit consumption in all three groups of children was similar at the start of the study, about one serving per day. The children who took part in either of the interventions increased their fruit consumption significantly compared to the control group, eating about two to two and a half servings per day. At one year follow up, only those children whose teacher had modeled fruit consumption continued to eat significantly more fruit, about two servings. Those who had completed the educational curriculum had decreased consumption back to the one serving a day of the control group.

Let me digress for a moment. When I was the same age as these study participants, my teacher, Miss Phillips, did something remarkable. She brought in fresh strawberries to the class and we made strawberry shortcakes.

In the half century since she did this, strawberry shortcake has occupied a special place in my life and I cannot recall turning down the opportunity to partake at any poine since. It wasn’t until I sat to write about this study that the association between my teacher’s cooking lesson and my subsequent craving for strawberry shortcake was appreciated.

It is remarkable the influence which the behaviors that teachers model can have on children, and in particular the persistence of their effect. This contrasts with the temporary impact of what they teach.

This brings to mind the weak effect seen in the PREDIMED Study that attempted to increase adherence to a Mediterranean style diet via five years of quarterly sessions with dieticians. This trial was the subject of an earlier commentary that I have written.

Neither children nor adults seem to change eating habits for long as a result of what they are taught. This study suggests they may change behavior based on what they see their role models do. In the case of children, based on what their teachers do. In the case of adults, perhaps they may mimic the behaviors their family members, friends or perhaps their doctors model for them.

My colleague Reiner Kremer confided to me that he keeps a Vita-Mix machine at his office and serves his patients freshly made smoothies in the waiting room. His may be a better strategy to improve a patient’s nutrition than all the lectures we might give. Perhaps keeping fresh fruit on our desks (and perhaps a bowl of nuts as well) will trigger a greater shift in behavior than our telling them to eat more fruit and nuts.

We may have greater success in modeling behavior than in lecturing about it. (Obviously even though we may be their doctors we may not be quite the role model their elementary school teachers were…...)

Eating fruit may not be the only healthful behavior that teachers can model for their students. Pondering the implications of this study raises the question whether schools should put less effort into curriculum development and more into teacher development, encouraging the teachers to model behaviors to their students that will increase their health and productivity throughout their lives?


Reference: Perikkou A, Gavrieli A, Kougioufa MM, Tzirkali M, Yannakoulia M. A novel approach for increasing fruit consumption in children. J Acad Nutr Diet. 2013 Sep;113(9):1188-93. doi: 10.1016/j.jand.2013.05.024. Epub 2013 Jul 16.