Parental concerns about picky eating and undereating, feeding practices, and child’s weight

Abstract

Objective:

Parents’ concerns about their child’s feeding may lead parents to pressure their child to eat, which may lead to a greater risk for obesity. We aimed to assess if parental concerns for picky eating and undereating are associated with pressure to eat and increased child BMI z-score (BMIz).

Methods:

We performed a cross-sectional study of 328 parents of healthy preschoolers assessing parent concerns about picky eating (Child Eating Behavior Questionnaire) and child undereating (“Are you concerned …doesn’t eat enough?”), parent pressure to eat (Child Feeding Questionnaire), and covariates. Dyads’ heights and weights were measured. Structural equation modeling (SEM) was performed to examine the relationships between parental concerns, pressure to eat, and child BMIz. Measurement models were tested and refined, and the structural model was tested. Model fit was determined using multiple goodness-of-fit indices.

Results:

Dyads were racially and socioeconomically diverse. The SEM model demonstrated good goodness-of-fit. Children who were perceived as not eating enough had significantly higher picky eating scores (β 0.756; p < 0.001). Parents had higher pressure to eat scores if children were more picky (β 0.148; p = 0.02) or were perceived as not eating enough (β 0.654; p < 0.001). Parental pressure to eat was not associated with the child’s BMIz.

Conclusions:

In a cohort of diverse preschoolers, parent concerns about eating were associated with increased pressure to eat, but pressure to eat was not associated with BMIz. Identifying these relationships is important to develop effective interventions to improve feeding practices in young children.

1. Introduction

Parents commonly express concern that their children are poor eaters [1,2]. As a feature of typical development, most children experience a reduction in appetite [3] and decreased rate of growth [4,5] during their body mass index (BMI) nadir. Food preferences are typically established during early childhood, though children’s preferences for certain foods may vary significantly, weekly or even daily [6,7]. Children who eat a “decreased variety of foods” are often considered to be picky eaters [8–10]. They commonly reject foods due to preferences in taste, texture, color, or food preparation, which is why this phenomenon is described on a spectrum. Childhood eating behaviors predict adult eating behaviors, [11,12] so it is important for primary care providers to help parents establish healthy eating behaviors early in life. Interestingly, parents often compensate for children’s pickiness by offering foods their children may find more acceptable, such as calorie-dense [13] or sweet [14] foods. It is postulated, but not well understood, that this parental response to picky eating may inadvertently increase the child’s risk for obesity [8].

Parental concerns about their child being a picky eater or not eating enough may change how parents feed their children, even among healthy weight children. Despite good intentions, this may contribute to negative feeding practices such as pressuring to eat. Pressure to eat is a feeding practice in which parents coax, bribe, or force their child to eat [15]. Parents often pressure their child to eat healthy foods, although this has been paradoxically associated with lower fruit and vegetable intake [9]. Parental pressure to eat is also associated with children’s decreased self-regulation of eating, [16] which places children at greater obesity risk [16]. While these relationships have often been examined in separate cohorts, it is unclear how parental concerns about child eating, parent feeding practices, child self-regulation of eating, and child weight status are associated. Therefore, among a sample of healthy 3- to 5-year-old children, we aimed to test our hypothesis that parental concerns for picky eating and undereating are associated with pressure to eat, self-regulation of eating, and increased child BMI z-score (BMIz).

2. Subjects, materials, and methods

2.1. Study overview

We performed a cross-sectional study including 328 parent-child dyads who attended a health care visit at one of four pediatric primary care clinics. The clinics included an urban clinic, an urban pediatric residency continuity clinic, a suburban clinic, and a rural clinic. After being roomed by nursing staff, potential participants were approached by research staff and informed about the study. After consenting to participate, parents completed a survey in the examination room on paper or on a tablet using REDCap. Parents were given the option to have the survey read to them to address low literacy, however, no parents chose this option. Parents were subsequently given a $10 gift card for their participation in the study. This study was approved by the Wake Forest School of Medicine Institutional Review Board.

2.2. Participants

Subjects were included if the parent/legal guardian (subsequently referred to as parent) was present at the visit, was ≥ 18 years old, and could read and write in English; and if the child was aged 3–5. Children were excluded if they were born prematurely (<37 weeks gestation); had a birth weight < 2500 g; had a chronic medical problem affecting weight gain patterns or prompting special diets (e.g. congenital heart disease, renal disease, food allergies); or had a documented feeding/eating disorder, developmental delay, or intellectual disability (including autism). Only one child per family was eligible for enrollment; if two eligible siblings were in clinic together, the research team approached the parent about participation for the younger sibling.

2.3. Measures

We assessed picky eating with the 6-item Food Fussiness subscale of the Child Eating Behavior Questionnaire (CEBQ) [17]. The Food Fussiness subscale is the most commonly used measure [1] and has validity and good internal consistency (Cronbach’s α = 0.88 in validation study, 0.89 in this sample) in low-income populations of young children in the United States [18]. Parents respond on a 5-point scale (never = 1, always = 5). Three questions (“My child enjoys tasting new foods,” “My child enjoys a wide variety of foods,” and “My child is interested in tasting food s/he hasn’t tasted before”) are reverse scored (never = 5, always = 1). Responses are averaged (range 1–5), with higher scores indicating higher picky eating. We assessed pressure to eat via the Child Feeding Questionnaire (CFQ). The pressure to eat subscale (4 items) of the CFQ has moderate internal consistency (Cronbach α = 0.7 in validation study, 0.63 in this sample) [15] and is validated in Black and Hispanic populations [19]. Parents respond on a 5-point scale and responses are averaged (range 1–5 with higher scores indicating higher pressure to eat). Child self-regulation of eating was assessed using three subscales of the CEBQ [20,21]. Satiety responsiveness, food responsiveness, and enjoyment of food are assessed on a 5-point scale (1 = never; 5 = always). The item “My child has a big appetite” (satiety responsiveness) was reverse scored (never = 5, always = 1). Higher scores on satiety responsiveness indicate higher self-regulation; higher scores on food responsiveness and enjoyment of food indicate lower self-regulation. These subscales are valid and have good internal consistency (Cronbach’s α = 0.70 for satiety responsiveness; 0.75 for enjoyment of food; 0.76 for food responsiveness in validation study; Cronbach’s α = 0.76 for satiety responsiveness; 0.91 for enjoyment of food; 0.84 for food responsiveness in this sample) in young Black and Hispanic children in the United States [22]. Parental concern for undereating was assessed using the question “Are you concerned that your child doesn’t eat enough?” with responses on a 5-point scale; responses are considered positive if parents agree or strongly agree [2].

As there is some evidence that parental feeding is influenced by parent mental health, we assessed various aspects of parental mental health and stress. Parental anxiety was assessed with the Generalized Anxiety Disorder-7 (GAD-7), a 7-item measure assessing anxiety over the past 2 weeks [23]. The GAD-7 has high internal consistency (Cronbach α = 0.92 in original study, 0.90 in this sample) and good test-retest reliability (intraclass correlation = 0.83). Parental depression was assessed using the Patient Health Questionnaire-9 (PHQ-9), a 9-item measure used to identify depression in the primary care setting (Cronbach α = 0.8 in validation study, 0.88 in this sample) [24]. Parental stress was assessed using the Perceived Stress Scale (PSS), a 10-item scale that assesses how unpredictable, uncontrollable, and overloaded participants find their lives. Responses are summed; a higher score indicates higher stress [25]. The PSS has high internal consistency (Cronbach α = 0.89 in original study, 0.82 in this sample) and test-retest reliability scores range from 0.55 to 0.85.

Demographic questions included parents’ income (<$20,000; $20-$39,999; $40-$59,999; $60,000 or more) and education (<high school; high school graduate; some college; associates, bachelors, masters, or higher). Parents also reported child’s sex, race and ethnicity (categorized in this analysis as Non-Hispanic Black, Non-Hispanic white, Hispanic, other), number of siblings, and age of siblings. We also assessed multiple social determinants of health, as these have previously been associated with parent feeding practices and child weight [26]. We examined household food insecurity using the 2-item screen by Hager et al. [27] An affirmative response to either question is a positive screen and has high sensitivity (97 %) and specificity (83 %) compared to the gold-standard 18-item United States Department of Agriculture Household Food Security Survey [28]. We used the Safe Environment for Every Kid Parent Questionnaire-R (SEEK PQ-R) Questionnaire to asses parental substance abuse, harsh punishment, and domestic violence [29]. We assessed neighborhood safety with the Built Environment Safety Scale. This is a 9-item measure assessing overall neighborhood safety, traffic, and crime that was designed for use in parents of preschool children [30]. Child anxiety and depression was assessed using the internalizing subscale of the Pediatric Symptom Checklist-17 (PSC-17), which has strong internal consistency (α = 0.91). Responses are summed; a score ≥ 5 indicates possible anxiety and/or depression. We included measures of mental health and social drivers of health as parents with mental health problems are more likely to report that their child experiences feeding problems [31] and social drivers of health have been associated with child feeding behaviors [32].

The child’s height, weight, date of visit, and date of birth was extracted from the child’s electronic medical record. Per standard clinical protocol, each child’s height and weight was measured by nursing staff using a wall-mounted stadiometer and mechanical beam scale. The parent’s height and weight was measured by research staff using the same wall-mounted stadiometer and mechanical beam scale. Clinical measurements have been shown to have good accuracy compared to height and weight obtained for research purposes [33]. Height and weight were used to calculate BMI. BMIz and BMI percentiles were then derived using the Centers for Disease Control and Prevention reference growth charts for age and sex [4]. Weight status was subsequently classified as underweight (BMI < 5th percentile for age and sex), healthy weight (5th to <85th percentile), overweight (85th to <95th percentile), and obesity (≥ 95th percentile).

2.4. Statistical analyses

We examined descriptive statistics from univariate analysis of categorical and continuous variables. Normality was assessed using histograms and the Shapiro-Wilk test. Unadjusted simple linear or logistic regression analyses were performed to examine bivariate associations of potential continuous and categorical covariates, respectively, with pressure to eat and child BMIz. Candidate covariates included child sex, age, and race/ethnicity; parent age, BMI, anxiety, depression, substance abuse, and stress; neighborhood safety; domestic violence; harsh punishment; household income and food insecurity; and family size. We included covariates in the structural equation modeling (SEM) model that were significant in the bivariate analysis with p < 0.1; this was done to avoid making the model more complex than needed and unnecessarily adding noise to the model.

We performed SEM to assess the relationships between parental concerns for their child’s weight, picky eating, undereating, and pressure to eat (Fig. 1, path diagram). As our SEM model has 32 parameters, using the rules of a minimum sample size of 200 subjects [34] or 10 subjects per parameter, [35] our sample size of 328 participants is sufficient for the SEM. Outcome variables included pressure to eat and BMIz. Predictor variables included picky eating and concern for undereating. Picky eating was conceptualized as a latent variable. Confounder variables of the pressure to eat and BMIz association included race/ethnicity, household income, household size, household food insecurity, harsh punishment, and parent depression. Measurement models for picky eating were tested and refined, and then the structural model was tested. Model fit was determined using multiple goodness-of-fit indices, including the Comparative Fit Index, Chi-Square Test of Model Fit, the Root Mean Square Error of Approximation (RMSEA), and the Standardized Root Mean Square Residual (SRMR), with values > 0.95 (CFI), ≥ 0.05 (Chi-Square), < 0.06 (RMSEA), and < 0.08 (SRMR) recognized as good model fit [36]. Data analysis was performed using Stata 14.0 and Mplus v7 software.

Fig. 1.

Structural Equation Modeling Path Diagram ^a,b. * denotes p < 0.05, *** p < 0.001. ^a Arrows in the path diagram do not represent causality due to the cross-sectional nature of this study. ^b Path coefficients represent estimate (standard error).

3. Results

3.1. Sample characteristics

Child participants (N = 328) were half female and were racially diverse with 39 % white, 40 % Black, and 17 % Hispanic. Two-thirds of children had a healthy weight while 14 % had overweight and 14 % had obesity (Table 1); mean BMIz z-score was 0.35 (SD 1.26). Picky eating and pressure to eat variables were normally distributed. Children had a mean (SD) picky eating score of 2.80 (0.80) and 13 % of parents expressed concern that their child did not eat enough. Parents self-reported a mean pressure to eat score of 2.71 (0.94) with 32 % of parents reporting high pressure to eat practices (Table 1).

Table 1.

Participant Characteristics.

	N (%) or Mean (SD) N = 328
Child Characteristic
Sex
Female	165 (50.3 %)
Male	163 (49.7 %)
Race/Ethnicity
White	126 (38.5 %)
Black	131 (40.1 %)
Hispanic	56 (17.1 %)
Other	14 (4.3 %)
Pediatric Symptom Checklist-17	6.37 (4.63)
BMI	16.4 (2.38)
BMI z-score [4]	0.35 (1.26)
Weight Status [4]
Underweight (<5th percentile)	18 (5.5 %)
Healthy Weight (5th to <85th percentile)	219 (67.2%)
Overweight (85th to <95th percentile)	44 (13.5 %)
Obesity (≥95th percentile)	45 (13.8 %)
Parent Characteristic
Relationship to Child
Mother	287 (87.5 %)
Father	36 (11.0 %)
Other	5 (1.5 %)
Education
<High School	55 (16.8 %)
High School Graduate	69 (21.0 %)
Some College	102 (31.1 %)
Associates, Bachelors, Masters, or +	102 (31.1 %)
Body Mass Index (BMI)^	31.3 (9.0)
Weight Status^
Underweight (BMI <18.5)	8 (2.5 %)
Healthy Weight (BMI 18.5–24.9)	73 (22.5 %)
Overweight (BMI 25.0–29.9)	83 (25.6 %)
Obesity (BMI ≥30.0)	160 (49.3 %)
Anxiety (Generalized Anxiety Disorder-7) [23]	3.52 (3.47)
Depression (Patient Health Questionnaire-9) [24]	3.8 (4.0)
Stress (Perceived Stress Scale) [25]	2.21 (0.67)
Harsh Punishment [29]	8 (2.44 %)
Domestic Violence [29]	17 (5.2 %)
Household Characteristics
Income
< $20,000	95 (29.1 %)
$20–40,000	86 (26.4 %)
$40–60,000	53 (16.3 %)
$60,000+	92 (28.2 %)
Household Food Insecurity [27]	32 (9.76 %)
Built Environment Safety Score [30]	8.35 (2.88)
Household Size
2–3 members	121 (36.1 %)
4 members	117 (34.9 %)
5+ members	97 (29.0 %)

^{^}N = 324.

Parent participants were mostly mothers (88 %); 17 % had less than a high school degree, 21 % were high school graduates, 31 % had some college, and 31 % had an associates, bachelors, masters, or higher degree (Table 1). Households had a wide range of incomes with 29 % making < $20,000, 26 % $20,000–40,000, 16 % $40,000–60,000 %, and 28 % making at least $60,000; 10 % of households were food insecure and the majority of households had at least 4 members (Table 1).

3.2. Bivariate associations of parental feeding concerns and pressure to eat

In unadjusted analyses, children who were perceived as more picky had 4.6 times the odds (95 % CI 2.8, 7.8) of also reporting that their child did not eat enough. Pressure to eat was significantly associated with both picky eating (β 0.2; 95 % CI 0.11, 0.30) and concern for undereating (β 0.1; 95 % CI 0.07, 0.15). Pressure to eat was not associated with parent BMI.

3.3. Bivariate associations between parental feeding concerns, pressure to eat, and child BMIz

Pressure to eat was associated with decreased child BMIz (β–0.08; 95 % CI −0.15, −0.002) but picky eating was not associated with child BMIz (β 0.07; 95 % CI −0.09, 0.24). Child BMIz was associated with parent BMI (β 0.03; 95 % CI 0.01, 0.04).

3.4. Structural equation modeling analysis

In the measurement model, the latent variable of picky eating was best assessed with the first four items of the Food Fussiness subscale of the CEBQ. For the measurement model, the model resulted in a comparative fit index of 0.99, a Chi-Square p-value of 0.99, a RMSEA of < 0.001, and a SRMR of 0.001 indicating good model fit.

The final SEM results are shown in Fig. 1. Picky eating was best assessed with 4 items of the food fussiness subscale. Picky eating was significantly associated with both parent concern for undereating and parental pressure to eat. Parent concern for undereating was also associated with parental pressure to eat. Pressure to eat was not associated with child BMIz. The final SEM model included six covariates that confounded the pressure to eat and BMIz association, including child race/ethnicity, household income, household size, household food insecurity, harsh punishment, and parent depression. In examining overall goodness of fit for the SEM model, the model resulted in a comparative fit index of 0.98, a Chi-Square p-value of 0.16, a RMSEA of 0.03, and a SRMR of 0.04, indicating good model fit. Children who were perceived as not eating enough had significantly higher picky eating scores (β 0.756; SE 0.144, p < 0.001). Parents had higher pressure to eat scores if children were more picky (β 0.148; 0.065, p = 0.02) or were perceived as not eating enough (β 0.654; 0.153, p < 0.001). Parental pressure to eat was not significantly associated with the child’s BMIz (Table 2).

Table 2.

Structural equation modeling results.

	Estimate	Standard Error	p-value
Picky Eating by:
Item 1	1.00	0	999
Item 2	1.22	0.68	0.07
Item 3	0.92	0.06	< 0.001
Item 4	0.85	0.07	< 0.001
Picky Eating on Concern for Undereating	0.76	0.14	< 0.001
Pressure to Eat on Picky Eating	0.15	0.07	0.02
Pressure to Eat on:
Concern for Undereating	0.65	0.15	< 0.001
Race/ethnicity	0.03	0.01	0.006
Household income	−0.02	0.01	0.02
Household Size	−0.08	0.04	0.03
Household food insecurity	0.28	0.16	0.09
Harsh Punishment	0.38	0.31	0.2
Parent Depression	0.003	0.005	0.4
Child BMIz on:
Pressure to Eat	0.64	0.47	0.2
Race/ethnicity	−0.06	0.10	0.6
Household income	0.03	0.07	0.7
Household Size	0.55	0.31	0.08
Household food insecurity	0.33	1.47	0.8
Harsh Punishment	−0.46	2.81	0.9
Parent Depression	−0.03	0.04	0.5

4. Discussion

In a sample of racially and socioeconomically diverse 3- to 5-year-old children, 13 % of parents were concerned that their children did not eat enough and 32 % of parents self-reported high pressure to eat feeding practices. Children who were perceived as not eating enough had significantly higher picky eating scores. Parents had higher pressure to eat scores if children were more picky or were perceived as not eating enough. Parental pressure to eat was not significantly associated with the child’s BMIz in the SEM model.

Our results that only 13 % of parents expressed concern that their child did not eat enough is lower than in our previous work, which showed that nearly a third of parents share this concern [37]. This previous study, which examined 286 female caregivers of low-income preschool children, used semi-structured interviews to assess concern for undereating, and the question varied slightly: “Do you ever worry that your child doesn’t or might not eat enough?” [37] Similar to our study, however, they showed that maternal concern for undereating was associated with both parent report of child picky eating and greater parental pressure to eat. Subsequently throughout the interview, mothers reported that their concern about undereating was rooted in a desire for healthy nutrition and adequate growth for their children. This parental desire for their child to have appropriate nutrition and growth likely drives the higher parental pressure to eat, as parents perceive it as their responsibility to ensure that their child’s diet is healthy with consumption of an appropriate amount of the right foods. In a separate study, we evaluated a different measure of picky eating that was composed of 3 factors: trying new foods, eating a sufficient quantity of foods, and desire for a specific preparation of foods [38]. Our findings suggest that for some parents, their perception of whether their child eats enough is actually an important component of whether they also perceive their child as picky. Additionally, consuming a sufficient quantity of food was significantly associated with parental pressure to eat, comparable to the results of this study.

A much greater body of literature exists on the associations of picky eating and pressure to eat. Many studies have demonstrated a positive relationship between child picky eating and parental pressure to eat [39,40]. A recent systematic review found that nonresponsive feeding practices, in which parents do not respond to their child’s hunger and fullness cues such as pressuring or restrictive feeding, were positively associated with picky eating [41]. It has been suggested that a bidirectional relationship is likely; parents may increase pressuring feeding practices in response to picky eating, and children who are pressured may have reduced enjoyment of eating and increased pickiness [42].

We found that pressure to eat was not associated with preschool-aged children’s weight status. Most cross-sectional studies have reported similar results, with either no association or that pressure to eat was negatively associated with children’s BMIz [43]. This is likely because in this instance, the pressure to eat is a response to their child’s body weight (assumed to be low) or eating behaviors. There are far fewer longitudinal studies examining this relationship, however. One recent study found that young children of low-income Hispanic mothers who are pressured to eat at ages 4–5 years old had a greater increase in BMIz after 3.5 years than children who were not pressured to eat [44]. This supports the theory that over time, consistent pressure to eat will lead to children overeating and increased weight gain. Additional longitudinal research is needed to further investigate these relationships through the course of childhood in diverse study populations.

Pediatric providers should ask about and directly address parents’ concerns about whether their child is a picky eater or eats enough, and discourage parents from pressuring their child to eat. Pressuring children to eat may diminish the child’s ability to recognize and respond to their feelings of hunger and satiety, and over time may lead to overeating. Parents should instead be encouraged to provide their child with healthy meals in a structured way (e.g. eating at the table together as a family and eating at consistent times), model healthy eating practices, and allow the child to choose how much of which foods they eat [45].

Strengths of this study include in-clinic assessments of child and parent body weight, along with over-recruitment of participants of Black race and those with a household income < $40,000, adding important data to the existing literature in parent feeding practices. Additionally, our study was strengthened by our inclusion of measures of multiple social determinants of health and parent and child mental health. However, these results should be considered within the context of the study limitations. This study was performed at pediatric primary care clinics associated with a single institution, and within a single geographic region in North Carolina. Respondents were all English-speaking and mostly mothers, which limits generalizability to similar populations. The data is cross-sectional, so inferences about causality cannot be determined and there may be a time effect on BMIz that was not captured. It is also possible that there is reverse causality, such that parents of children with obesity may restrict their child’s eating and have lower pressuring feeding practices. Future longitudinal work should assess parent concerns about feeding, pressure, restriction, and child BMIz throughout childhood. Parental concern for undereating was measured using one item, and although this method has been used previously, [2] it is possible that this question could be interpreted in different ways by parents. Finally, all data was based on parental self-report, so it is possible that parent concerns about their child’s feeding are at least partially due to parent perceptions or misperceptions about their child rather than an objective and quantifiable behavior. We also did not observe parental feeding practices directly, but many studies have shown consistency between parent-reported and observed feeding practices [37,44].

4.1. Conclusion

In a sample of racially and socioeconomically diverse 3- to 5-year-old children, parent concerns about whether their child ate enough or was a picky eater were associated with increased pressure to eat. Identifying the relationship of parental concerns and how these concerns impact feeding practices is essential to developing effective pediatric obesity prevention interventions to improve feeding practices and establish healthy weight trajectories in young children. Pediatric providers should be prepared to address the frequent concerns that parents of young children have about their child’s eating habits and should offer guidance on healthy feeding practices. Additional research, including observing parent feeding practices to further explore parent influence on children’s food consumption, appears warranted. There is also the need to follow parental concerns and feeding practices along with child weight throughout childhood in diverse cohorts to elucidate parental influence on a child’s future eating habits and weight.