Clinical use of a food inventory to identify maternal underreport on children's food intake: experience of a reference center in Brazil

Background: Feeding Difficulties (FD) are a common problem in childhood, and dietary assessments are extremely important to evaluate food consumption and selectivity patterns. In clinical practice, it is often observed that caregivers have an exacerbated perception of the severity of their children's FD and this may impact on dietary reports. Objectives: To compare the variety of foods consumed by children with FD under maternal perception to the evaluation performed by nutritionist using a food inventory in association with traditional methods. Methods: It is a cross sectional study with 119 mother children pairs, with complaints of FD followed at an outpatient service in Sao Paulo, Brazil. Data collected consisted of demographics, maternal caregiving style, responsibility over feeding the child, perception of children's feeding diversity (expressed in numbers, taken from a self-explanatory food inventory), presence of organic disease, type of FD, BMI Z-score, and actual repertoire of foods consumed (expressed in numbers, assessed by the nutritionist after evaluating the food inventory). The inventory provided information about the food items the child accepts without rejection, used to accept but now rejects, and completely rejects. Spearman correlation and Lynn's coefficient of variation, ANOVA and T-student tests were used, with a 5% significance level. Results and discussion: Overall food diversity ranged from 16 (p25%) to 30 (p75%) different types of foods. Maternal perceived food diversity ranged from 4,3 to 14,5 (p25-75). Around 23,7% of children were considered highly selective (less than 15 types of foods) (p=0,000). Maternal perception underestimated that assessed by the nutritionist in 2,2 times, with low reproducibility between these two variables (r=0,14). Both maternal perception and professional assessment did not vary according to age, organic comorbidities, gender, maternal parity, feeding style or responsibility over feeding routines (p>0,49). Findings enhance the importance of parental behavioral interventions and of the use of dietary research methodologies which are complementary to already validated strategies. Conclusions: There was maternal underestimation on the quantity of foods accepted by children with FD. The data reinforce the need for family orientation about the expectations regarding the feeding practices and preferences of their children.