Parental perception of child's weight, their attitudes towards child's dietary habits and the risk of obesity

Perception is the construction of mental symbols or representations of reality acquired through the senses [1], and because it can be influenced by various factors, perception is not always faithful to reality. Child’s age, sex and family history of obesity are the factors most often found to be associated with a parental misperception of child’s weight [2‐7]. Child’s body mass index (BMI) deserves a special mention. Most of the evidence on the relationship between parental perception of weight and actual child’s BMI z scores comes from cross-sectional studies [5, 6]; therefore, the direction of that association and whether it represents a true biological effect remain unclear.

The high prevalence of parental underestimation of child’s weight status [4] has important public health implications. Evidence on the association of weight misperception with lifestyle [3, 8, 9] and feeding practices [10, 11] is scarce and inconsistent; therefore, the extent to which misperception represents an obstacle to obesity prevention is unknown [12].

Attitudes are emotional, motivational, perceptive, and cognitive beliefs that influence the behavior or practice of an individual regardless of whether he/she has knowledge [13]. Far from being static and fixed, attitudes are continuously shaped by intrinsic (genetic, age, and sex) and environmental factors, such as family, friends, or neighborhood [14]. Parents are important agents in the promotion of the health of their children. Parents play a key role in structuring their children’s first experiences with food and eating through their own beliefs, food practices, perspectives, eating attitudes, dietary knowledge, and understanding of the benefits of food and nutrients on health [15]. Parental nutrition knowledge and dietary attitudes have been described as particularly important factors for children’s healthy food knowledge [15].

To our knowledge, the association between parental perception of child’s weight and parental attitudes towards child’s dietary habits has not been studied. In this scenario, we aimed to assess the association between parental underestimation of child’s weight and parental attitudes towards child’s dietary habits using data from the SENDO project (from Spanish Child’s follow-up for an optimal growth), an ongoing cohort of Spanish children.

At recruitment, children's mean (SD) age was 5.1 (0.8) years, and 347 (51%) of the children were boys. Of the 674 participants (including 596 clusters, i.e., different families) who had completed the baseline questionnaire when this study started, 111 (16%) had parents who underestimated their weight and 67 (10%) had parents who overestimated it.

Participant’s main characteristics by the parental perception of child’s weight are presented in Table 1. Mothers were the person who most often completed the questionnaire in all 3 groups. No differences were observed between groups for maternal age, the percentage of mothers with higher education, family history of obesity or the number of children. Parents in the three groups reported similar knowledge on nutritional recommendations for children. Regarding children’s characteristics, no differences were observed for age and sex distribution, but regarding race, those who underestimated it were less often white (91.89% vs. 97.38%; P = 0.005).

Regarding lifestyle variables, underestimation was not associated with energy intake, physical activity or snacking. A marginally significant difference (P = 0.19) was observed for adherence to the Mediterranean dietary pattern, with lower adherence in children whose parents underestimated their weight as compared with those who correctly perceived it. Children whose parents underestimated their weight were more likely to watch television during meals (P = 0.04) and spent more screen-time (P = 0.06). On the other hand, parents who overestimated their child’s weight reported a longer duration of the breastfeeding (P = 0.004), less physical activity (diff. = − 9.48, METs-h/week; 95% CI [-14.99 to -3.97]) and less screen-time (diff. = − 0.24 h/day; 95%CI [0.04–0.44]) than those who correctly perceived child’s weight.

The main analyses included 607 children (111 in the underestimation group and 496 in the correct estimation group) in 543 clusters (i.e., different families). After accounting for sex and race, parental underestimation of the child’s weight was associated with significantly higher odds of unhealthy attitudes towards child’s dietary habits [OR (95% CI): 3.15 (1.60–6.18)] (Table 2). Similar estimates were obtained when we progressively adjusted for lifestyle variables OR (95% CI) for unhealthy attitudes: 2.88 (1.48–5.60) and child’s z score of BMI [OR (95% CI)] for unhealthy attitudes: 3.35 (1.71–6.53) (Table 2). The estimates (95%) for the other factors analyzed in the multivariable adjusted are presented in Supplementary table1.

In this cross-sectional study of 674 preschoolers from the SENDO project, we found that parents who underestimated their child’s weight had 3.35 -fold higher odds (95% CI: 1.71–6.53) of having unhealthy attitudes towards child’s dietary habits. We previously published that parental attitudes towards child’s dietary habits were inversely associated with child’s nutritional adequacy and diet quality [27], but as far as we know, the association between parental perception of child’s weight and parental attitudes towards child’s dietary habits has previously not been reported. Therefore, our findings are of great interest because they may help pediatricians and public health practitioners identify a subgroup of patients who are most likely to have unhealthy attitudes towards their child’s dietary habits.

Pediatricians may play an important role in parental perception of healthy weight and their preferences regarding obesity counseling [28]. Correct perception of child’s weight is thought to make parents more willing to adopt favorable changes in lifestyle [29], but a recently published systematic review [5] concluded that interventions aiming to correct that misperception may be harmful. We agree with the idea that preventing and treating childhood obesity should be a specialist’s main concern regardless of the perception of the parents [5]. However, our results suggest that asking about the parental perception of their child’s weight may be helpful by identifying a potential group at risk. Efforts should then be directed to identify each unhealthy habit that may exist and to help the family correct these habits individually, progressively and gradually.

Sixteen percent of parents in the SENDO project underestimated their child’s weight. This prevalence is much lower than that reported in previous studies [4], especially considering that information was collected with a verbal scale, which is more prone to error than a visual scale [30]. This difference may be attributed to participants’ idiosyncrasies because most participants in the SENDO project have highly educated parents.

Beyond the scale used, several reasons have been proposed to explain why some parents tend to underestimate their child’s weight status. First, they may be reluctant to acknowledge that their child is overweight or obese because of the social stigma attached to it [31]. Second, the increased prevalence of childhood obesity may alter their perspective, so that when they perceived their child’s weight is similar to that of a large part of the population of his/her age, they may incorrectly conclude it is normal[3]. Related to this alteration of perspective, family history of obesity [4, 7] also has been suggested to be associated with a parental misperception of child’s weight. However, our results did not support this hypothesis. Finally, the image that media offers of obese children is often that of extreme cases of morbid obesity, which may lead to a normalization of cases of overweight or slight obesity [4].

We also observed a strong association between child’s BMI z score and parental perception of child’s weight. This finding agrees with previous studies which suggested that child’s z score of BMI was the strongest predictor of parental underestimation. Importantly, most of the evidence suggesting a positive association between child’s z score of the BMI and the odds of parental underestimation of child’s weight comes from cross-sectional studies, wherein reverse-causality bias cannot be discharged [6, 32]. Prospective studies assessing the association between parental perception of child’s weight and weight gain are scarce, and these studies are not exempt from limitations, such as suboptimal control of confounding [31, 33] or inconsistency in the definition of the exposure [31, 34]. Other factors that have been suggested to be associated with parental underestimation of child’s weight are sex and age. Our results do not support previous evidence probably because of the low variability in participant’s age and because differences in fat distribution and secondary sexual characters between boys and girls are non-existent at 5 years old, which was the mean age of the participants in our study.

Fully understanding the relation between parental perception of child’s weight and his/her risk of obesity is very complex because the association between parental underestimation of child’s weight and his/her prospective risk of obesity is confounded with child’s actual weight (Fig. 1). In this scenario, our study is of great importance because parental attitudes towards child’s dietary habits could be a mediator in this association. Previous studies showed that dietary habits are associated with weight change [35] and cardiovascular disease [36] in adult populations. Nevertheless, the hypothesis that unhealthy dietary attitudes towards one’s child’s dietary habits lead to a higher risk of the child be overweight or obese needs to be confirmed in prospective studies.

Some limitations of our study must be acknowledged. First, owing to the cross-sectional design of the study, causality cannot be inferred. Cross-sectional studies are susceptible to reverse-causality bias, and therefore we cannot completely reject the association between weight misperception and parental attitudes towards child dietary habits as being in the opposite direction or even bidirectional. Second, we used self-reported information, which is susceptible to misclassification bias. Nevertheless, a validation study of the anthropometric measures reported by parents of participants in the SENDO project showed high correlation coefficients between the measured and the reported data [37]. Third, our sample was not representative of the pediatric population. Participants in the SENDO project are mostly white children with highly educated parents from a developed European country. Although this factor may hamper the generalizability of our results, it also has some benefits, such as higher validity of the self-reported information and reduction of potential confounding by socio-economic variables [38]. However, we acknowledge that the homogeneity of the sample, with little prevalence of overweight or obesity, may have affected our results and may have led to wide confidence intervals. Lastly, although the questionnaire on parental attitudes towards child’s dietary habits has not been validated, previous studies on adult populations showed that it is associated with diet quality, weight change and cardiovascular disease [35, 36, 39].

Despite these limitations, our study has several strengths. Our findings add to the existing evidence by providing data which suggest that parental perception of child’s weight is associated with their attitudes towards child’s dietary habits. Our findings thus suggest a link between parental misperception of child’s weight and his/her future risk of overweight or obesity. In addition, the baseline questionnaire of the SENDO project is extensive and allowed for the control of several confounders.

In conclusion, we found that parental underestimation of child’s weight was associated with 3.35-fold higher odds (95% CI: 1.71 – 6.53) of unhealthy attitudes towards child’s dietary habits, which we hypothesized may be the link between parental misperception and prospective weight gain. Although the relationship between parental perception, child’s actual weight status and dietary attitudes may be complex, we believe that pediatricians and public health practitioners should pay attention to the parental perception of their child’s weight to identify parents who underestimate it as an at-risk group in which to inquire about their lifestyle and dietary habits.