This pilot study adds to the emerging data on the effectiveness of specific anticipatory guidance. The mothers in both intervention groups (MOMS and OP) had healthier child feeding habits than those in the BF control group, with the maternal-focused MOMS group performing well in the number of servings of fruits and vegetables given to the infants, in spite of having no direct instructions to do so. Additionally, the OP group of mothers reported fewer family meals than MOMS and BF (usual care) by age 12 months.
There are positive differences in what mothers feed their infants following simple advice to reduce obesogenic maternal-eating habits, such as snacking instead of meals, television watching while eating and caloric beverage and fast food consumption, when compared with mothers who did not receive such advice. Although there has been a virtual storm of publication about childhood obesity in the past 5 years, there have been few attempts to study interventions and fewer still looking at prevention interventions. Those that have been published have generally taken place in schools and other institutions where there is some level of control of the subjects.
This is the first publication of an intervention in infancy in the pediatric office. Our choice to work with maternal eating patterns was made as we realized that the eating patterns of children are closely related to that of their families. Recent correlational work by Anderson and Whitaker25
would suggest that the recommendations of eating scheduled meals with other family members and turning off the television might be particularly helpful. The ecology of family eating is one that is built on over time. The difficulty of changing behaviors after they are well established has been demonstrated by our difficulties in treating obesity. The approach of improving maternal eating habits allows us to attempt to improve the eating environment without intruding into the feeding of the infant with specific discussions of serving size and running the risk of interfering with infant satiety signals. It also has the potential advantage of avoiding the intercultural miscommunications that may occur with the recommendation of specific foods.
This study has some very clear limitations. Because of the risk of intervention contamination had we tried to randomize at the patient level, we randomized at the clinic level. In spite of our attempts to select clinics with similar demographics, the samples had differences in race, marital status, and educational level, which may explain some of the baseline differences in maternal eating habits. Although the differences we found remained even after controlling for baseline differences, it is certainly possible that these differences are accompanied by others that were undetected and affected the outcome, making this a major weakness in the study. Variables that differed among the groups and variables that are known to relate to infant obesity, such as breastfeeding were controlled for, but owing to small sample sizes, the model could not be expanded to include all of the variables that were collected. We did not correct for use of the Supplemental Nutrition Program for Women, Infants, and Children, food stamps, and public insurance. This adjustment would not have influenced the study results owing to the high prevalence of participation in all of these programs. As a pilot, it was done in one community and it may not generalize to other communities. Finally, the study is primarily a study of self-report. Although the items came from validated tools, it is possible that the mothers were telling us what we wanted to hear; however, it is interesting that infant feeding was not part of the instructions given to the MOMS mothers, yet the group had significant differences in reported infant intake.
Although it would have been gratifying to demonstrate differences in weight among groups, it is not surprising that at this early age such a difference would not be found. This was a short trial that did not go past 12 months. The next step may be to carry out both intervention and evaluation into the toddler years when children are even more integrated into family eating style.
In spite of the apparent positive effects of the intervention, there is much room for improvement in all groups. For instance, none of the groups came close to the recommendations of the Centers for Disease Control and Prevention for fruit and vegetable servings per day. Although the MOMS group had the lowest juice intake, the 12-month-olds were still drinking more than 8 oz of juice a day and 23% were watching more than 2 hours of television per day with 39% of homes having the television on for more than 8 hours per day. It is also worth noting that the interventions may have had unintended effects. For example, the OP group reported eating fewer family meals than the BF and MOMS groups at 12 months. This may be related to some unmeasured baseline difference in the groups, or may be that the very specific portion size suggestions disrupted family meals because of the need to monitor what the infant ate more closely.