Maternal Eating Styles and Restrictive Feeding Practices: Indirect Effects through Perceived Child Appetite and Weight Concern

1. Introduction

Childhood obesity and unhealthy eating behaviors remain pressing public health concerns globally and in Chile. According to the World Health Organization (WHO), more than 39 million children under the age of five were overweight or obese in 2020, with prevalence rates steadily increasing across both developed and developing nations [1]. Chile, in particular, ranks among the countries most affected in South America [2]. According to national monitoring data from the Junta Nacional de Auxilio Escolar y Becas (JUNAEB) [3], approximately half of Chilean children entering kindergarten are classified as overweight or obese. These alarming statistics highlight the urgent need to identify psychosocial factors contributing to early eating behaviors and weight-related outcomes.

Parental influence is among the strongest determinants of children’s dietary habits, food preferences, and weight outcomes [4]. The home food environment and parental feeding practices shape not only what children eat but also how they develop attitudes toward food and intake self-regulation [5]. Feeding strategies such as restriction, pressure to eat, and monitoring have been consistently linked with children’s eating patterns and risk of obesity [6,7]. Importantly, beyond specific feeding practices, parental attitudes and dispositions toward child feeding play a crucial role. In particular, parents who are highly concerned about their child’s weight may engage in restrictive feeding to prevent weight gain, yet such strategies can paradoxically increase children’s preference for restricted foods and reduce their ability to regulate appetite [8,9].

A growing body of research has emphasized that parents’ own eating behaviors and psychological tendencies are integral to understanding how they approach child feeding [10,11]. Maternal eating styles—specifically restrained, emotional, and external eating—have been conceptualized as dispositional tendencies that influence how mothers perceive and respond to food-related cues [12,13]. Mothers high in restrained eating may be particularly vigilant about their children’s food intake, while those high in emotional or external eating may misinterpret children’s hunger or satiety cues [14,15]. Such tendencies shape maternal subsequent feeding practices [16,17].

Another key factor in this dynamic is how mothers perceive their children’s eating attitudes and appetitive traits. The Children’s Eating Behaviour Questionnaire (CEBQ) [18] captures parental perceptions of traits such as food responsiveness, enjoyment of food, and satiety responsiveness. These perceptions may or may not reflect children’s actual eating behaviors but stands to reason they likely guide how parents monitor and regulate their children’s feeding. For instance, perceiving a child as highly food responsive (i.e., having a strong drive to eat) may prompt greater use of restrictive feeding practices [17].

Although previous research has examined components of the pathway linking parental and child eating-related variables, few studies have tested comprehensive models integrating maternal eating styles, perceptions of children’s appetitive traits, concern about child weight, and restrictive feeding. For example, Rodgers et al. [16] found that specific maternal feeding practices—particularly instrumental and emotional feeding—predicted subsequent increases in children’s obesogenic eating behaviors and BMI, while maternal monitoring was associated with decreases in food-approach behaviors. However, they did not consider how mothers’ own eating styles might shape their perceptions of children’s appetitive traits or their restrictive feeding. In contrast, Pickard et al. [17] demonstrated that parents’ own eating profiles were directly linked to their children’s latent eating profiles, with these associations mediated by feeding practices such as using food for emotional regulation and maintaining a healthy food environment. Yet, this study did not incorporate parents’ perceptions of children’s appetitive traits or weight-related concerns. Complementary evidence from Haycraft and Blissett [15] showed that higher parental BMI was associated with more controlling feeding, suggesting that parents’ own weight and eating tendencies influence mealtime control. Finally, Russell et al. [17] provided longitudinal evidence that parental restriction and pressure predicted greater food responsiveness and lower satiety responsiveness in children—and that children’s eating traits, in turn, evoked more parental restriction—highlighting reciprocal dynamics but not the motivational roles of parental concern or perception.

Taken together, these studies show that maternal eating styles, perceptions of children’s appetitive traits, and concerns about child weight each contribute to parental feeding practices. However, existing work has examined these links in isolation or within partial models, leaving untested an integrated, sequential model connecting maternal eating styles, perceptions, concern, and restrictive feeding. Addressing this gap is particularly relevant in the preschool years (ages 2–6), when children are highly dependent on caregivers for food provision and when foundational self-regulation patterns in eating are established.

Study Objectives and Hypotheses

The present study aimed to test a structural equation model examining whether maternal eating styles predict restrictive feeding practices indirectly through mothers’ perceptions of their children’s appetite and their concern about child weight. Based on previous findings, we hypothesized that:

1. Higher levels of maternal restrained, emotional, and external eating would predict higher perceived child drive to eat (CEBQ-derived latent factor).

2. Greater perceived child drive would predict higher maternal concern about child weight.

3. Both perceived child drive and maternal concern would, in turn, predict more restrictive feeding practices.

4. The relationship between maternal eating styles and restrictive feeding would be primarily indirect, operating sequentially through perceived child drive and maternal concern.

2. Materials and Methods

2.1. Participants

Participants were 488 mothers of children aged 2 to 7 years. Mothers’ mean age was 33.87 years (SD = 4.81; range = 20–49), and their mean body mass index (BMI) was 26.95 kg/m² (SD = 5.52; range = 15.92–52.07), placing the sample, on average, in the overweight category according to World Health Organization (WHO) criteria. Mothers reported an average of 1.81 children (SD = 0.95; range = 1.72–7.47). Children’s mean age was 3.85 years (SD = 1.33).

2.2. Procedure

Participants were recruited online via targeted advertisements and posts on social media platforms (e.g., Facebook, Instagram) directed to Chilean parenting communities. Eligibility criteria required participants to be mothers of at least one child aged 2–7 years and to be primarily responsible for feeding that child. After providing electronic informed consent, mothers completed a self-administered questionnaire hosted on a secure online platform. To encourage participation, respondents were entered into a raffle for one of 30 gift cards valued at 45,000 Chilean pesos (CLP), about 45 US dollars.

2.3. Ethics

The study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of the School of Psychology, Universidad Adolfo Ibáñez. All participants provided informed consent prior to participation.

2.4. Measures

Maternal Eating Styles. The Dutch Eating Behavior Questionnaire (DEBQ) [12] was used to assess three maternal eating style dimensions: Restrained Eating (Restr_E; 10 items; e.g., “Do you deliberately eat less in order not to become heavier?”; α = .86), Emotional Eating (Em_E; 13 items; e.g., “Do you have a desire to eat when you are irritated?”; α = .93), and External Eating (Ext_E; 10 items; e.g., “If food smells and looks good, do you eat more than usual?”; α = .84). Responses were rated on a 5-point Likert scale (1 = “never” to 5 = “very often”). These subscales demonstrated strong internal consistency and factorial validity in Chilean and Latin American samples. In this study, the three subscale scores were used as indicators of the latent construct Maternal Eating Styles (M_ED_x).

Child Feeding Attitudes and Practices. The Child Feeding Questionnaire (CFQ) [6,7] was employed to assess mothers’ perceptions of child weight, concerns about child weight, and feeding-related practices. Four subscales were used: Concern about Child Weight (Cn_cw; 3 items; e.g., “How concerned are you about your child eating too much when you are not around?”; α = .79), Restriction (Restrict; 8 items; e.g., “I intentionally keep some foods out of my child’s reach”; α = .82), Monitoring (Monit; 3 items; e.g., “How much do you keep track of the high-fat foods your child eats?”; α = .75); and Prohibition (Prohib; 3 items; e.g., “I forbid my child from eating certain foods”; α = .77). In the SEM, Maternal Worry (M_worry) was defined by the Concern items, and Maternal Restriction (M_Restricc) was defined by the Restriction, Monitoring, and Prohibition global scores.

Perceived Child Eating Attitudes. The Children’s Eating Behaviour Questionnaire (CEBQ) [18] assesses parental perceptions of children’s appetitive traits across eight subscales. The present study focused on the food pro-ingestive dimension, which integrates the subscales Food Responsiveness (F_Resp; 5 items; e.g., “My child is always asking for food”; α = .84), Emotional Overeating (Em_Eat; 4 items; e.g., “My child eats more when worried”; α = .76), and Enjoyment of Food (Enjoy_F; 4 items; e.g., “My child loves food”; α = .88). Responses were given on a 5-point Likert scale (1 = “never” to 5 = “always”). These subscales jointly defined the latent variable Child Drive (Ch_drive), representing maternal perceptions of children’s appetite and motivation to eat.

All measures were validated Spanish-language versions with strong psychometric properties and demonstrated reliability in Chilean and Latin American populations.

2.5. Data Analysis

Structural equation modeling (SEM) analyses were conducted in R using the lavaan package [20]. Models were estimated using maximum likelihood with robust standard errors (MLR), which provides Huber–White robust SEs and a Satorra–Bentler (SB)–scaled χ² test of model fit [21]. Model adequacy was evaluated using robust comparative fit indices (CFI, TLI), the robust RMSEA with 90% confidence intervals, and the SRMR. AIC and BIC were examined for model parsimony.

Four latent variables were specified:

• Maternal eating styles (M_ED_x): DEBQ Restrained, Emotional, and External eating total factor scores.
• Child drive (Ch_drive): CEBQ Food Responsiveness, Enjoyment of Food, and Emotional Overeating total factor scores
• Maternal worry (M_worry): The three CFQ concern-factor items.
• Maternal restriction (M_Restricc): CFQ Restriction composite, Monitoring, and Prohibition factor-total scores.

Two nested models were tested:

Full model: Included direct paths from maternal eating styles to child drive, maternal worry, and maternal restriction, as well as indirect paths through the mediators (Figure 1).

Reduced model: Identical measurement structure but omitting nonsignificant direct paths from maternal eating styles to maternal worry and restriction. Indirect and total effects (e.g., a₁c₃ and a₁b₃c₄) were computed within lavaan using labeled parameters and tested for significance with the delta method under MLR. Model comparison used the SB-scaled χ² difference test (lavTestLRT) and inspection of robust fit indices and AIC/BIC values.

3. Results

Model Fit

The reduced model provided a good to excellent fit to the data, χ²(50) = 163.60, p < .001, CFI = .94, TLI = .92, RMSEA = .07, SRMR = .06 [22]. Model comparison indicated that removing the two nonsignificant direct paths (M_ED_x → M_worry_over, M_ED_x → M_Restricc) did not degrade fit, Δχ²(2) = 1.07, p = .59, and slightly improved AIC (29,711 vs. 29,714) and BIC (29,829 vs. 29,840). See Table 1 for complete fit indices.

Structural Paths Standardized path coefficients for the reduced model are shown in Table 2. Maternal eating styles significantly predicted greater child drive (β = .44, p < .001). In turn, child drive predicted both higher maternal worry (β = .39, p < .001) and higher maternal restriction (β = .28, p < .001). Maternal worry also strongly predicted restriction (β = .65, p < .001). Indirect effects indicated two significant mediated pathways linking maternal eating styles to maternal restriction: (a) through child drive (β = .12, p < .001) and (b) sequentially through child drive → maternal worry (β = .11, p < .001). The combined indirect effect was significant (β = .23, p < .001). The total effect of maternal eating styles on restriction remained significant (β = .17, p = .008), even though direct effects were nonsignificant.

4. Discussion

This study tested an integrative model linking maternal eating styles to restrictive feeding practices through mothers’ perceptions of their children’s appetitive traits and concern about child weight in a Chilean sample. Structural equation modeling revealed that maternal restrained, emotional, and external eating were indirectly associated with restrictive feeding via child drive and maternal worry, whereas direct effects were nonsignificant. These findings highlight the role of maternal perceptions and cognitions as mechanisms through which parental traits shape feeding control.

4.1. Interpretation and Theoretical Implications

Consistent with our hypotheses, mothers who reported greater restrained, emotional, or external eating perceived their children as having stronger pro-ingestive drives (e.g., food responsiveness, enjoyment of food, and emotional overeating). This association supports the idea that parents’ own eating tendencies color their perceptions of their children’s appetite cues, a pattern previously observed by Temmen et al. [24]. Both studies suggest that maternal eating tendencies, whether emotional or externally driven, influence how parents interpret and respond to their children’s hunger and satiety signals—often amplifying perceived risk of overeating. These biases likely arise through modeling or projection, whereby parents’ own preoccupations with food and weight heighten their sensitivity to, or misinterpretation of, children’s interest in food [15].

Perceived child drive, in turn, predicted both heightened maternal concern about weight and greater restrictive feeding, consistent with prior evidence that parental perceptions of high appetite or weight risk trigger controlling feeding practices [8]. The strong association between maternal concern and restriction underscores the proximal role of worry about child weight gain in eliciting restrictive control. Collectively, these patterns align with cognitive-behavioral models proposing that parental cognitions—such as perceived appetite risk and weight concern—mediate the link between parental eating tendencies and feeding practices [18]. The absence of a direct path from maternal eating styles to restriction further suggests that interventions should target perceptual and cognitive processes rather than focusing exclusively on behavioral regulation of feeding

4.2. Relation to Previous Research

Previous studies have examined individual components of this pathway but rarely within a single integrated framework. For example, parents’ perceptions of their children’s weight, rather than actual BMI, has been found to predict restrictive and pressuring practices [9]. Similarly, parents’ concern about weight and diet quality appears to predict their use of controlling strategies, particularly among overweight children [8]. Studies of parental eating patterns also support indirect influences: Temmen et al. [24] showed that maternal emotional and external eating predicted greater “feeding to soothe” during infancy, which in turn increased infant food responsiveness. Likewise, Singh et al. [23] reported that maternal restrained eating was associated with more pressure feeding, while emotional and external eating interacted with child weight to predict contradictory concerns about under- or overeating. These findings, together with those from Pickard et al. [19], who identified parallel parent–child eating profiles mediated by feeding practices, converge to suggest that parental eating styles influence feeding behavior indirectly through perceptions and cognitions.

The present study advances this literature by modeling both perceptual (child drive) and cognitive (weight concern) mediators simultaneously and by confirming their sequential contribution to restrictive feeding. This approach integrates prior variable-centered and person-centered findings into a single, theoretically coherent framework, extending transactional models of feeding and appetite regulation [16].

4.3. Practical and Clinical Implications

The findings have important implications for family-based interventions. Efforts to promote adaptive feeding practices should focus not only on observable behaviors but also on the parental cognitions that drive them. Helping parents recognize normal variability in children’s appetite and avoid pathologizing robust eating could reduce unnecessary restriction and meal-related tension. Psychoeducational programs might also address maternal emotional and external eating patterns that amplify weight-related worry, incorporating cognitive-behavioral or mindfulness-based strategies to recalibrate parents’ interpretations of children’s eating behavior. By targeting perceptual accuracy and emotional regulation, such interventions could disrupt a self-reinforcing loop linking maternal anxiety, child feeding control, and children’s subsequent risk of dysregulated eating.

4.4. Strengths, Limitations, and Future Directions

This study contributes novel evidence from a large, non-WEIRD Latin American sample, addressing calls to diversify the populations represented in behavioral research [25,26]. Additional strengths include validated Spanish-language measures and a theoretically grounded SEM model that captures the sequential interplay among eating styles, perceptions, concern, and restriction. However, limitations include the cross-sectional design, which precludes causal inference, and reliance on self-report data, which may introduce shared-method bias. Recruitment via online platforms may also limit generalizability to less connected populations or fathers.

Future research should employ longitudinal and observational methods to clarify temporal ordering and bidirectionality between child appetitive traits and parental feeding. Incorporating fathers and multiple caregivers would further illuminate gender and contextual differences in these dynamics.

4.5. Summary

In summary, the findings support a model in which maternal eating styles shape restrictive feeding indirectly through their influence on perceptions of children’s appetite and weight-related concern. Mothers exhibiting higher restrained, emotional, or external eating are more likely to view their children as food-driven, to worry about potential weight gain, and to impose restrictive control. Addressing these perceptual and cognitive mechanisms may be key to developing interventions that foster more balanced, responsive feeding dynamics and healthier eating development during early childhood.