The role of parental health knowledge and practices in mitigating obesity risks among preschool children

Abstract

Some studies are concerned about parental Knowledge and behaviours regarding their kids’ obesity; however, few studies introduce a comprehensive framework of preschool children concerning parental health knowledge, attitude, and practice involving socioeconomic factors for predicting body mass index and body fat percentage. In our research, we considered families that only had one preschooler. Two hundred fifty-six families with a girl and 176 families with a boy participated in the survey conducted among both groups. One family with two children was not considered for this study. We considered structural equation modelling involving mediating analysis. The Body Mass Index (BMI) z-score and the Body Fat Percentage (BFP) were employed in this study to represent childhood obesity. The results showed that parental socioeconomic status has a positive relationship with the parents’ KAP. Still, it has a negative relationship with the child’s BMI z-score and the child’s BFP. Furthermore, the effects were more pronounced in families that had a boy. The Parents’ KAP also mediates the relationship between the P-SES and the child’s BMI z-score and BFP. Furthermore, the relationships were found to be stronger among families that had a boy. This study highlights the significant impact of parental health knowledge, attitudes, and practices on the prevalence of obesity in preschool-aged children. The findings underline parents’ essential role in shaping their children’s health habits and outcomes, implying that focused interventions to educate and empower parents could be critical in preventing childhood obesity. Effective health education programs and supportive policies that raise parental Knowledge and encourage healthy habits can lead to significant advances in preventing and treating obesity in early childhood, promoting healthier generations to come.

Introduction

The prevalence of overweight children with obesity is increasing at an alarming rate, rendering juvenile obesity a significant public health issue globally. In the near term, this disease may result in several health issues, including type 2 diabetes, hypertension, and joint complications. Over time, it may also result in severe chronic illnesses such as cardiovascular disease, type 2 diabetes, and certain cancers¹. Multiple variables contribute to childhood overweight and obesity. These elements encompass genetic, behavioural, and environmental influences that interact intricately². The primary causes include poor dietary practices, insufficient physical activity, and social and environmental influences. Children are increasingly becoming overweight due to dietary shifts toward high-calorie, low-nutrient meals, reduced outdoor playtime due to urban living and more screen time^3,4. There has been a significant increase in childhood obesity in China, particularly among preschoolers⁵. This increase parallels a global trend and reflects China’s distinct social, cultural, and economic transformations. Accelerated urbanisation, economic expansion, and alterations in living arrangements and familial structures have led individuals to consume more Western-style fast food and sugary beverages while simultaneously exhibiting reduced physical activity^6,7. Until recently, legislation mandated that each family could have just one child. The regulation resulted in the solitary child frequently receiving the entirety of attention and resources from their parents and grandparents, thus leading to overfeeding and diminished physical activity. Moreover, a societal conviction exists in China that a well-nourished infant signifies health and affluence, exacerbating the issue further⁸. To combat childhood obesity in China, the following specific factors must be considered. The evidence indicates that public health initiatives must be culturally attuned and tackle the fundamental socioeconomic and lifestyle determinants contributing to this escalating epidemic. Parents’ Knowledge, attitudes, and behaviours toward health are pivotal in effectively managing their children’s body weight, as they directly influence their offspring’s dietary habits and physical activity. Informed parents are more likely to understand the importance of balanced nutrition and the dangers of excessive calorie intake, enabling them to make informed decisions about the food and portion sizes they offer. They can foster a healthful home environment by encouraging physical activity and limiting sedentary behaviours, which can influence health and fitness perceptions. Elliott⁹ posited that parents who adopt a nutritious diet and consistent exercise habits act as exemplars for their children, establishing a basis for healthy living inside the home. Behavioural modelling is crucial, as children often replicate the tendencies exhibited by their primary caregivers.

Knowledge, attitude, and practices (KAP) framework

The KAP survey, also known as the KABP [Knowledge, Attitudes, Beliefs, and Practices] survey, is a methodological tool initially created to study family planning practices globally by Ratcliffe¹⁰. The KAP framework is a sociological paradigm that evaluates how an individual’s Knowledge, attitudes, and behaviours (practices) regarding a subject might impact results in many sectors such as public health, education, environmental conservation, and marketing. The framework is highly beneficial for creating, executing, and assessing treatments and programs focused on modifying habits or enhancing results in a specific demographic¹¹.

Knowledge refers to the information individuals possess regarding a specific subject or topic. It includes the Knowledge, data, and information that individuals or groups have. Knowledge may involve comprehending a disease’s aetiology and preventive strategies in public health. Attitudes refers to individuals’ feelings about a particular topic and inclinations toward specific actions or customs. Attitudes can be favourable or unfavourable and are shaped by personal beliefs, cultural conventions, and societal influences. They are essential in influencing an individual’s likelihood of adopting or altering new practices. Practices refer to the behaviours individuals engage in based on their knowledge and attitudes. This may encompass habits, rituals, and specific behaviours that are developed gradually.

Gaps of previous studies

Although a considerable body of literature has investigated the current childhood obesity epidemic, especially modifiable risk factors, some critical areas are still underexplored, especially the influence of parents. The first step in comprehending childhood obesity is to analyse parental knowledge, attitudes, and practices (P-KAP) and parental socioeconomic status (P-SES), which can contribute to children’s health behaviours.

Lack of research on P-KAP in the Asian context and preschool students

In recent decades, the KAP survey has emerged as an essential tool in public health research and program evaluation, offering significant insights into individuals’ health-seeking behaviours and the efficacy of public health interventions^12,13. KAP surveys assess individuals’ Knowledge, attitudes, and practices concerning health issues, facilitating the identification of gaps in awareness, misconceptions, and behavioural impediments within communities¹⁴. By comprehending these elements, public health experts can formulate customised actions, guaranteeing that health programs adequately meet unique requirements and enhance long-term health results. KAP surveys are crucial for the design, execution, and ongoing assessment of health initiatives, enabling policymakers and healthcare practitioners to enhance resource allocation and augment the effectiveness of public health endeavours.

Investigations of the KAP framework have been undertaken across many locations, encompassing America¹⁵, Canada¹⁶, the Middle East¹⁷, and Europe¹⁸. Nonetheless, a notable research deficiency persists in examining how P-KAP affect childhood obesity, especially within the Asian context. Straughan and Xu¹⁹ emphasised that few studies have particularly investigated parental Knowledge, attitudes, and behaviours concerning obesity in Asia, creating a significant gap in comprehending the impact of parental influence on children’s health habits. Moreover, although research has examined the direct impact of P-SES on childhood obesity, insufficient focus has been directed towards how P-SES affects childhood obesity via P-KAP, especially among Chinese parents^20,21.

No research to date has particularly utilised the KAP framework for preschool-aged children in this setting, despite early childhood being a critical developmental phase where parental influence on diet, physical activity, and overall health habits is most significant. Preschool-aged children predominantly depend on parental guidance for their dietary selections and lifestyle practices, rendering P-KAP a crucial factor in the risk of childhood obesity. The deficiency of research constrains the capacity of health professionals and policymakers to formulate effective, evidence-based programs to prevent childhood obesity from an early age.

Limitations of BMI as the sole indicator of obesity

Researchers frequently employ BMI as a principal metric for assessing obesity levels owing to its simplicity, cost-efficiency, and ease of computation. BMI evaluates height and weight to broadly evaluate body mass, applicable to persons of all ages and genders, with modifications for children based on age and sex percentiles. Nevertheless, BMI alone fails to distinguish between adipose tissue and lean tissue, which may result in misclassification, such as categorising muscular persons as overweight or overlooking extra body fat in those with a normal BMI.

BFP is an adjunct measure to mitigate this restriction, providing a more accurate body composition assessment by quantifying fat mass relative to total body weight. In contrast to BMI, BFP offers information on fat distribution, differentiating between subcutaneous and visceral fat, each with distinct metabolic and health consequences. Visceral fat increases the risk of metabolic diseases, including diabetes and cardiovascular disease.

Employing BMI and BFP concurrently improves obesity evaluation by distinguishing various obesity phenotypes, including persons with normal BMI yet elevated body fat (normal-weight obesity) and those with high BMI but reduced body fat (muscular build). This dual strategy allows researchers and healthcare practitioners to create more tailored obesity interventions, ensuring patients receive suitable counsel based on their real body composition rather than weight alone.

Lack of understanding of the mediating role of P-KAP in the relationship between P-SES and childhood obesity

Current studies often investigate the direct effects of P-KAP and P-SES on children’s BMI Z-SCORE, emphasising the independent influence of parental behaviours and socioeconomic factors on obesity prevalence. Nevertheless, there is a paucity of comprehension regarding the indirect mechanisms by which P-SES influences children’s BMI z-score through P-KAP, resulting in a significant void in the literature. P-SES is essential in determining parents’ access to health-related resources, Knowledge, and capacity to adopt healthy practices. Parents possessing elevated educational attainment and financial security are more inclined to obtain dependable health information, nutritious foods, organised physical activities, and high-quality healthcare services. Conversely, low-SES parents may encounter financial, educational, and environmental limitations that hinder their capacity to obtain or implement sufficient Knowledge regarding good diet, portion control, and physical activity guidelines. This knowledge gap may result in suboptimal feeding patterns, characterised by an overdependence on processed or high-calorie foods due to their price and convenience.

Furthermore, parental attitudes toward health and nutrition, influenced by socioeconomic conditions, affect the degree to which they prioritise healthy habits in the household. Parents from elevated socioeconomic status may regard obesity prevention as a crucial component of child development, prompting them to implement organised routines for food preparation, physical activity, and diminished screen time. In contrast, parents of lower socioeconomic status may possess an inadequate understanding or alternative conceptions of the risks associated with childhood obesity, thus underestimating the long-term repercussions of bad dietary practices and sedentary lifestyles. Parental behaviours, encompassing feeding patterns, food preparation routines, and activity encouragement, are the behavioural connection between parental socioeconomic status and childhood obesity outcomes. Families possessing superior financial and educational resources are more inclined to embrace health-promoting practices, such as preparing nutritionally diverse home-cooked meals and promoting organised physical activity. Conversely, low-SES families may contend with time limitations, food shortages, or insufficient safe play areas, resulting in increased dependence on fast food and screen-based entertainment, which contribute to childhood obesity. This study underscores the mediating function of P-KAP in the association between P-SES and childhood obesity, stressing that interventions must address not just socioeconomic disparities but also Knowledge, attitudes, and behavioural modifications. Addressing these mediating factors can offer a more holistic strategy for diminishing childhood obesity rates, especially among families of lower socioeconomic status.

This study

This study expands current literature by presenting a coherent conceptual framework delineating the links among P-SES, P-KAP, and childhood obesity. P-SES, which includes factors such as income, education, and occupation, affects the degree to which parents can obtain and implement health-related Knowledge and behaviours. P-KAP is a pivotal mediator that converts socioeconomic advantages (or disadvantages) into health consequences for children.

To enhance the theoretical foundation of this investigation, we utilise the Stimulus-Organism-Response (SOR) theory, which offers a systematic framework for comprehending how external stimuli influence internal cognitive and behavioural mechanisms, resulting in measurable outcomes. As per SOR theory, an external environmental component (stimulus) affects an individual’s internal state (organism), thereby prompting behavioural reactions²². This study applies the framework to address childhood obesity issues in preschool children in China, conceptualising P-SES as the stimulus (S), P-KAP as the organism (O), and children’s BMI z-score and BFP as the reaction (R).

In this scenario, P-SES is the external determinant of parents’ access to health-related Knowledge, attitudes, and practices. Parents from elevated socioeconomic situations typically possess superior education, financial security, and access to health-related information, influencing their comprehension and methodology regarding nutrition, physical activity, and overall child health practices. P-KAP denotes parents’ intrinsic cognitive and behavioural disposition, illustrating how their understanding and perspectives on health manifest in tangible activities, including food preparation, promotion of physical activity, and healthcare choices. Ultimately, children’s BMI z-score and BFP represent the definitive health outcomes influenced by parental factors stemming from their socioeconomic status.

By including SOR theory in this study, we offer a more thorough theoretical elucidation of how parental socioeconomic position affects childhood obesity via parental cognition and behaviour. This theoretical framework elucidates the mechanisms connecting parental socioeconomic conditions to childhood health outcomes and provides valuable insights for developing targeted interventions. The principal objective of this study is to investigate the mediating function of P-KAP in the association between P-SES and childhood obesity, thereby providing a more refined comprehension of how parental factors influence children’s health. This study enhances theoretical frameworks and practical implementations in developing treatments to mitigate socioeconomic inequities in childhood obesity prevention initiatives.

The current study intends to investigate the following:

1. The relationship of P-SES with parents’ (a) knowledge, (b) attitude, and (c) practices among boys’ and girls’ preschool children.

2. The relationship of parents’ Knowledge with (a) child’s BMI z-score and (b) child’s BFP among boys’ and girls’ preschool children.

3. The relationship of parents’ attitudes with (a) child’s BMI z-score and (b) child’s BFP among boys’ and girls’ preschool children.

4. The relationship of parents’ practices with (a) child’s BMI z-score and (b) child’s BFP among boys’ and girls’ preschool children.

5. The mediating roles of parents’ Knowledge, attitudes, and practices between P-SES and the child’s BMI z-score.

6. The mediating roles of parents’ Knowledge, attitudes, and practices between P-SES and children’s BFP.

The proposed model is depicted in Fig. 1.

Fig. 1.

Research framework.

Materials and methods

Measures

In SEM, measuring the study variables correctly is essential so that the model is correct and easy to understand. The first step is to use a few observed indicators or measurement variables to operationalise latent variables. Latent variables are theoretical constructs that cannot be directly observed. The operationalisation is based on theoretical Knowledge and previous research. This ensures that the indicators used are true and accurate ways to show the hidden constructs. There are four latent variables and two measurement variables in this study. Latent variables included P-SES, parents’ Knowledge, parents’ attitudes, and parents’ practices. Measurement variables included BMI z-score and BFP.

This study incorporated three latent variables into the P-KAP framework: parents’ Knowledge, attitudes, and practices. Straughan and Xu¹⁹ developed a questionnaire to assess P-KAP, a measure used to estimate childhood obesity. The questionnaire comprises nine questions, including three items designed to assess parents’ Knowledge, three designed to measure parents’ attitudes, and three designed to measure parents’ practices. The questionnaire utilised a seven-point Likert scale to score all KAP issues. The scaled from 1 (indicating “Strongly Disagree”) to 7 (indicating “Strongly Agree”). P-SES included five measurement variables. These are the mother’s age, the mother’s education, the father’s education, the father’s age, and the family income.

Childhood obesity is assessed using the BMI z-score (BMI-for-age z-score) derived from the WHO growth standards, which calibrate BMI based on a child’s age and sex. The BMI z-score reflects a child’s BMI relative to a reference population, a valuable tool for evaluating growth and nutritional status. This assessment is utilised due to its widespread recognition and frequent application in prior studies evaluating childhood obesity^23,24. Per WHO guidelines, a BMI z-score below − 2.0 is categorised as underweight, whilst a score ranging from − 2.0 to + 1.0 is deemed normal weight. Children with a BMI z-score ranging from + 1.0 to + 2.0 are designated as overweight, while those with a BMI z-score over + 2.0 are classed as obese. This standardised methodology has been commonly employed in obesity research to enable health practitioners to monitor growth trajectories, discern concerns for obesity-related health complications, and execute early interventions. BMI z-scores, which consider age and sex variations, provide a more dependable assessment of childhood obesity than raw BMI readings alone.

Sampling process

Pilot study and sample size

Urumqi City, situated in Xinjiang Province, contains 32 preschools with enrolments surpassing one hundred children. A multi-stage sampling procedure was utilised to guarantee a representative sample. The initial phase entailed cluster sampling, with each preschool regarded as its cluster. This approach facilitated the selection of schools that offered a varied representation of preschool children from diverse regions. Initially, all 32 preschools were contacted through email and telephone to evaluate their willingness to participate in the study. Nonetheless, owing to administrative limitations and scheduling issues, only seven preschools affirmed their willingness to cooperate. A pilot study was done before the main data collection to ascertain the reliability and validity of the survey instrument. A total of 50 surveys were disseminated to parents of preschool children from various institutions. Out of them, 46 responses were deemed legitimate, yielding a substantial response rate of 92%. The initial analysis of the pilot study validated that the measuring constructs were dependable and appropriate for application in the comprehensive investigation.

A power analysis was performed using G*Power statistical software²⁵ to ascertain the suitable sample size for the primary investigation. G*Power is a commonly utilised instrument for determining the necessary sample size based on statistical factors, ensuring that the study has adequate power to identify significant effects.

This study performed a power analysis for multiple linear regression to examine the links among parental Knowledge, attitudes, practices, and children’s obesity outcomes. The subsequent statistical parameters were employed in the analysis:

Alpha level (α) = 0.05 (standard threshold for statistical significance).

Power (1−β) = 0.80 (recommended minimum criterion for identifying genuine effects).

Effect size (f²) = 0.15 (medium effect size as suggested by Kraft²⁶.

Using these parameters, G*Power calculations indicated that a minimum sample size of 396 participants was required to achieve adequate statistical power. This ensured that the study had an 80% probability of detecting an actual effect if one existed while maintaining a 5% probability of committing a Type I error (false positive).

Collecting data

A total of 560 questionnaires were delivered, with 80 surveys allocated to each of the seven participating preschools. The data collection method was conducted closely with school principals and teachers, who were instrumental in encouraging parental participation. Parents were enlisted during preschool events and ceremonies to guarantee a high response rate, where they were already on the school premises. This method enabled parents to complete the survey on-site while interacting with school personnel and researchers. Alongside completing the questionnaire, parents were asked to submit their children’s health cards to acquire precise health-related information, including height, weight, and immunisation records. This measure reduced mistakes linked to self-reported health information, augmenting the gathered data’s reliability.

Each questionnaire required around 15–20 min to complete, and parents were allowed to seek clarification from researchers or school staff if needed. Informed consent was secured from all involved parents before data collection commenced. Participants were thoroughly apprised of the study’s objectives, methodologies, and entitlements, and a formal consent document was supplied, explicitly indicating that participation was voluntary and that they might withdraw at any moment without repercussions. Parents were solicited in several ways to generate a broad and representative sample of participants. School principals and teachers were instrumental in dispatching formal invites to parents via school communication tools, newsletters, and direct communications. Moreover, in-person invitations were issued during school events and ceremonies, where parents were already in attendance and could engage. These events provide an accessible environment for parents to interact with the researcher and complete the survey on-site. Moreover, educators and school personnel promoted involvement by elucidating the study’s significance in comprehending childhood obesity and parental impacts on health behaviours.

Furthermore, confidentiality and anonymity were rigorously upheld throughout the investigation. The data gathered from parents was kept confidential, and replies were anonymised whenever feasible to safeguard their privacy. The researcher guaranteed that the information supplied by participants remained untraceable to individual respondents. Before initiating data collection, the study secured ethical approval from the Ethics Committee to ensure compliance with ethical research standards and pertinent guidelines for research involving human participants. Upon completing the data collection process, 432 out of the 560 distributed questionnaires were deemed valid and included in the final analysis. The remaining 128 responses were excluded due to incompleteness, inconsistencies in the provided data, or lack of essential information. This rigorous screening method ensured the incorporation of exclusively high-quality data in the study, augmenting the results’ validity and reliability.

Inclusion and exclusion criteria

The following were the inclusion criteria for the study:

• Mothers or primary caregivers of preschool children (aged 3–6 years old) were identified from one of the included preschools.

• The child possessed a current health card with accurate weight, height, and immunisation status records.

• Parents could both read and understand the language of the questionnaire and the consent form.

• Single-child families were also chosen to prevent the clustering effect and intra-family bias in parental response.

The exclusion criteria included:

• Families with more than one preschooler attend the same preschool facility. The aim was to have statistically independent observations and avoid possible bias if the parent wrote from a generic viewpoint that would apply to more than one child.

• Children with known comorbid chronic diseases or illnesses related to growth (e.g., endocrine diseases or congenital diseases), which could interfere with the BMI estimation.

• People with incomplete or inconsistent questionnaire answers or missing health card information.

Results

Descriptive statistics

Table 1 presents the descriptive statistics of P-SES.

Table 1.

Descriptive statistics of P-SES.

Variable	Mother (number)	Mother (percentage)	Father (number)	Father (percentage)
Age of parents (years)
Less than 25 years	102	23.6%	85	19.7%
26–30	101	23.4%	85	19.7%
31–35	120	27.8%	97	22.5%
36–40	72	16.7%	112	25.9%
More than 40 years	37	8.6%	53	12.3%
Income of parents (RMB)
Less than 3000	78	18.1%	30	6.9%
3000–5000	156	36.1%	78	18.1%
5001–8000	154	35.6%	128	29.6%
8001–10,000	32	7.4%	109	25.2%
More than 10,000	12	2.8%	87	20.1%
Education level of parents
Less than high school	18	4.2%	19	4.4%
High School	56	13.0%	67	15.5%
Diploma	178	41.2%	202	46.8%
Bachelor	128	29.6%	120	27.8%
Master or PhD	52	12.0%	24	5.6%
Job experience of parents (years)
Less than 5 years	102	23.6%	34	7.9%
5–10	187	43.3%	228	52.8%
11–15	103	23.8%	134	31.0%
16–20	33	7.6%	32	7.4%
More than 20 years	7	1.6%	4	0.9%

The age group between 31 and 35 accounts for the most significant percentage of mothers (27.8%), whereas the age group between 36 and 40 accounts for the most significant percentage of fathers (25.9%). This information is derived from Table 1. Between RMB5000 and RMB8000, the highest percentage of individual income for both moms and fathers was between 35.6% and 29.6% for mothers and fathers, respectively. Most mothers and fathers had a high school diploma (41.2%; 46.7%), and the occupations with the highest proportion of years of experience for both mothers and fathers were between five and ten years (43.3%; 52.8%).

Validity, reliability, and multicollinearity analysis

To assess the validity and reliability of a survey employing Structural Equation Modelling (SEM) analysis, Fornell and Larcker²⁷ delineate a series of requisite conditions that must be fulfilled. These prerequisites guarantee that the assessed constructs are both conceptually valid and statistically reliable. A principal metric of internal consistency reliability is Cronbach’s alpha (α), which must possess a coefficient of 0.7 or greater for a latent variable to be deemed reliable. This threshold signifies that the survey items related to a specific construct demonstrate sufficient internal consistency. As seen in Table 2, all Cronbach’s alpha values surpass the required threshold of 0.7, thus affirming the reliability of the constructs employed in this investigation. Alongside Cronbach’s alpha, construct validity is evaluated using convergent validity, commonly quantified by the Average Variance Extracted (AVE). Segars²⁸ asserts that the AVE value must exceed 0.5 to indicate that the latent variable accounts for more than half of the variance of its indicators. An AVE exceeding 0.5 validates that the items within a construct possess a substantial share of their variance, reinforcing the model’s validity. According to Table 2, all AVE values meet or surpass the threshold, demonstrating that the constructs possess adequate convergent validity. In addition to reliability and validity, it is essential to assess multicollinearity before evaluating the structural model. Elevated collinearity among independent variables might skew regression estimates and diminish model interpretability. Hair, Black²⁹ proposes that Variance Inflation Factor (VIF) values under 5 signify an acceptable degree of multicollinearity. This analysis revealed that all VIF values were within the allowed range (< 5), indicating that multicollinearity is not a serious concern in the model. By satisfying the Cronbach’s alpha, AVE, and VIF criteria, the study guarantees that the measurement model is dependable and valid, facilitating precise interpretation of the structural relationships within the SEM framework.

Table 2.

Validity, reliability, and multicollinearity analysis.

Variables	Cronbach Alpha	AVE	VIF
Parents with a boy
P-SES	0.709	0.513	[3.11, 4.23]
Parents’ knowledge	0.763	0.513	[2.76, 3.24]
Parents’ attitude	0.765	0.587	[3.12, 3.92]
Parents’ practices	0.811	0.601	[3.47, 4.67]
Parents with a girl
P-SES	0.711	0.521	[2.67, 3.88]
Parents’ knowledge	0.725	0.546	[3.34, 4.56]
Parents’ attitude	0.733	0.531	[2.76, 3,69]
Parents’ practices	0.709	0.509	[3.19, 3.78]

Model fitting

In SEM, the model’s fit is one of the most important ways to judge whether or not a research model is suitable and valid. According to Schermelleh-Engel, Moosbrugger³⁰, a good model must have a fit index greater than 0.9, meaning it accurately describes the data. The Goodness of Fit Index (GFI), the Relative Fit Index (RFI), the Comparative Fit Index (CFI), the Tucker-Lewis Index (TLI), and the Normed Fit Index (NFI) are some of the most important indices for judging model fit. Furthermore, the Root Mean Square Error of Approximation (RMSEA) must be below 0.05 to signify a well-fitting model. These indices measure how well the suggested model fits with the data that has been collected.

Several indices, such as GFI (0.776), RFI (0.721), CFI (0.876), and NFI (0.833), were below the 0.9 threshold, showing that the first model was not well fitted (See Fig. 2). Modification values were used to improve the model fit. All of the indices went above 0.9 after the changes were made. The GFI went up to 0.908, the RFI to 0.923, the CFI to 0.915, the TLI to 0.936, and the NFI to 0.911 (see Fig. 2). The initial RMSEA value was 0.059, indicating a decent fit; however, following the application of modification indices, it improved to 0.046, reflecting a superior approximation of the data. All these improvements show that changes, like releasing parameters or adding more relationships, effectively fixed problems in the model structure after the changes were made, the model “fit” well, which means that it correctly shows how the data is structured.

Fig. 2.

Model fit analysis.

Structural model

The results of structural equation modelling for male and female preschool-aged children are shown in Tables 3 and 4, respectively. Research has shown a significant positive connection between P-SES and the Knowledge, attitudes, and practices of parents of preschool-aged children, regardless of whether they are male or female. The Knowledge of the parents has a significant relationship with both the BMI z-score and the BFP of the child. As the second positive KAP characteristic, the parents’ attitudes had a substantial positive link with the child’s BMI z-score and BFP. Both the child’s BMI z-score and the child’s BFP are substantially connected with the practice of the parents.

Table 3.

Direct, indirect, and total effects (Dependent variable: child’s BMI z-score).

Relationships	Estimates	LL 95% CI	UL 95% CI
Family with a girl (N = 256)
Direct effect
P-SES→ Parents’ knowledge	0.515	0.414	0.584
P-SES→ Parents’ attitude	0.466	0.397	0.501
P-SES→ Parents’ practices	0.394	0.309	0.44
Parents’ knowledge→ Child’s BMI z-score	−0.546	−0.584	−0.505
Parents’ attitude→ Child’s BMI z-score	−0.397	−0.424	−0.319
Parents’ practices→ Child’s BMI z-score	−0.413	−0.469	−0.357
P-SES→ Child’s BMI	−0.428	−0.474	−0.376
Indirect effect
P-SES → Parents’ knowledge → Child’s BMI Z-score	−0.281	−0.311	−0.212
P-SES → Parents’ attitudes → Child’s BMI z-score	−0.185	−0.202	−0.145
P-SES → Parents’ practices → Child’s BMI Z-score	−0.163	−0.204	−0.119
Family with a boy (N = 176)
Direct effect
P-SES→ Parents’ knowledge	0.562	0.502	0.616
P-SES→ Parents’ attitude	0.494	0.412	0.54
P-SES→ Parents’ practices	0.441	0.415	0.482
Parents’ knowledge→ Child’s BMI z-score	−0.554	−0.557	−0.48
Parents’ attitude→ Child’s BMI z-score	−0.516	−0.567	−0.463
Parents’ practices→ Child’s BMI z-score	−0.661	−0.71	−0.605
P-SES→ Child’s BMI z-score	−0.482	−0.521	−0.413
Indirect effect
P-SES → Parents’ knowledge → Child’s BMI Z-score	−0.311	−0.367	−0.257
P-SES → Parents’ attitudes → Child’s BMI z-score	−0.255	−0.298	−0.196
P-SES → Parents’ practices → Child’s BMI Z-score	−0.292	−0.363	−0.239

*<0.05; **<0.01; ***<0.001

Table 4.

Direct, indirect, and total effects (Dependent variable: child’s BFP).

Relationships	Estimates	LL 95% CI	UL 95% CI
Family with a girl (N = 256)
Direct effect
P-SES→ Parents’ knowledge	0.51	0.463	0.561
P-SES→ Parents’ attitude	0.461	0.413	0.525
P-SES→ Parents’ practices	0.389	0.323	0.429
Parents’ knowledge→ Child’s BFP	−0.541	−0.582	−0.502
Parents’ attitude→ Child’s BFP	−0.392	−0.463	−0.335
Parents’ practices→ Child’s BFP	−0.408	−0.461	−0.363
P-SES→ Child’s BFP	−0.423	−0.479	−0.364
Indirect effect
P-SES → Parents’ knowledge → Child’s BFP	−0.276	−0.321	−0.233
P-SES → Parents’ attitudes → Child’s BFP	−0.181	−0.219	−0.139
P-SES → Parents’ practices → Child’s BFP	−0.159	−0.187	−0.131
Family with a boy (N = 176)
Direct effect
P-SES→ Parents’ knowledge	0.592	0.53	0.615
P-SES→ Parents’ attitude	0.489	0.455	0.53
P-SES→ Parents’ practices	0.436	0.412	0.479
Parents’ knowledge→ Child’s BFP	−0.59	−0.529	−0.429
Parents’ attitude→ Child’s BFP	−0.548	−0.592	−0.502
Parents’ practices→ Child’s BFP	−0.437	−0.47	−0.41
P-SES→ Child’s BFP	−0.442	−0.486	−0.4
Indirect effect
P-SES → Parents’ knowledge → Child’s BFP	−0.349	−0.381	−0.245
P-SES → Parents’ attitudes → Child’s BFP	−0.268	−0.336	−0.216
P-SES → Parents’ practices → Child’s BFP	−0.191	−0.265	−0.118

*<0.05; **<0.01; ***<0.001

The bootstrap estimation method was used to determine the significance of the partially mediated model. The bootstrap sample size chosen was 5,000. The effects of KAP features that act as mediators are illustrated in Tables 3 and 4, respectively. This table aims to highlight the effect that P-SES plays in influencing the BMI z-score of preschool-aged boys and girls. A comparison of P-SES’s influence on affecting a child’s BFP is presented in Table 4, which includes both preschool children and girls.

Discussion

The study aimed to examine the role of KAP in the connection between P-SES and childhood obesity in preschool-aged children. The study utilised SOR theory as the theoretical foundation. The study used P-SES as the stimulus, KAP as the organism, and BMI z-score and BFP as the responses to explain childhood obesity concerns in preschool children. The study aims to examine the impact of P-SES on the Knowledge, attitudes, and practices of preschool children in China. The study also attempted to investigate the impact of P-SES on both BMI z-score and BFP. The study also examined how Knowledge, attitudes, and practices mediate P-SES, BMI z-score, and BFP associations. Two hundred fifty-six girls and 176 boys from seven preschools in Xinjiang province, Urumqi, China, were included in the study. Online surveys were done to gather replies.

The study’s innovative contributions to theory extend multiple fields. The study contributes to the P-SES literature by identifying strong positive associations between P-SES and KAP in preschool boys and girls. The P-SES causes KAP in people³¹. The alterations in the P-SES also impacted the Knowledge, attitude, and conduct of families with preschool children³². The study aligns with the research conducted by Straughan and Xu¹⁹ and would benefit families with preschool children.

Second, this study provides important findings from the existing literature on knowledge, attitudes, and practices within families with preschool children. It reveals a notable inverse correlation between KAP and the children’s BMI z-score and BFP, consistent with the extant literature^33–35. The results indicate that families with greater Knowledge, favourable attitudes towards healthy practices, and superior food and physical activity habits are associated with children with lower BMI z-score and BFP. This association underscores the importance of parental understanding and practices in influencing early childhood health outcomes. The study convincingly shows how KAP directly influence key indices of child health, making a strong argument for interventions focused on enhancing health education and behaviours among parents. This may result in improved nutritional outcomes and healthier growth patterns in preschool-aged children, highlighting the significance of specific educational and behavioural approaches in promoting early childhood health.

The contribution of the study is to show a further similar mediation by parental KAPs of the relationship between P-SES and key health outcomes, including BMI z-score and BFP, among preschool children. This research indicates that the influence of P-SES on children’s health, particularly their BMI z-score and BFP, is channelled through parental knowledge, attitudes, and practices. While similar mediation pathways have been examined in other countries³⁶, this study contributes new insights within the context of China, where cultural norms, educational patterns, and socioeconomic disparities may uniquely shape parenting behaviours and their impact on early childhood health outcomes. The potential outcomes of this output are that individuals with a higher socioeconomic status typically have improved access to information, tools, and environments that promote healthy eating and physical activity³⁷. Parents of a better socioeconomic status may understand diet and physical activity more, have more favourable views on health and wellness, and have habits that promote a healthy lifestyle. These factors can directly impact their children’s food habits, physical activity, and overall health behaviours, affecting their BMI z-score or BFP^38,39. Lower socioeconomic status can restrict access to nutritious meals and safe exercise environments and decrease the chances of obtaining education on healthy behaviours. This can harm children’s BMI z-score or body fat percentage due to less educated or less healthy Knowledge, attitudes, and practices connected to nutrition and physical activity.

The correlation between P-SES and childhood health outcomes is well-documented; nonetheless, comprehending the precise processes by which P-SES affects P-KAP is crucial for developing successful interventions. Parents with elevated socioeconomic status frequently possess enhanced access to educational resources, healthcare facilities, and nutritious food selections, fostering improved nutritional awareness and favourable health behaviours. Moreover, parents of higher socioeconomic status are more inclined to participate in health-oriented decision-making, including promoting physical exercise, enrolling children in organised sports programs, and restricting screen time. Conversely, low-SES households may have financial and environmental limitations restricting access to nutritious food, secure recreational areas, and instructional resources regarding nutrition and physical activity. Stress and extended working hours in households with lower socioeconomic status may diminish parental involvement in healthy food preparation and promoting an active lifestyle. By identifying these mechanisms, interventions can be customised to address these disparities, including establishing community-based educational initiatives, subsidised nutritious meal programs, and policies that enhance access to recreational facilities for families of low socioeconomic status. Mitigating these structural impediments can improve P-KAP among parents of lower socioeconomic status, ultimately resulting in improved results for their offspring.

Comprehending the indirect effects of P-KAP on a child’s BMI (BMI z-score) offers essential insights for formulating effective obesity prevention measures. This implies that the study should not solely focus on direct variables, such as a parents’ wealth or education level, but also examine how these factors influence parental actions, which subsequently impact a child’s health outcomes. A parent with enhanced nutritional Knowledge is more inclined to offer balanced meals, control portion sizes, and restrict unhealthy snacks, all of which facilitate healthier weight management in children. Parents exhibiting favourable attitudes toward physical activity are more inclined to promote outdoor play, enrol their children in sports programs, or impose screen time restrictions, indirectly diminishing sedentary behaviours and fostering a healthy BMI. Conversely, parents lacking understanding or possessing detrimental attitudes may inadvertently perpetuate harmful eating behaviours, such as regular fast-food consumption, resulting in significant weight increases in children. By acknowledging these indirect routes, practitioners can formulate focused interventions, such as parental education programs or behavioural coaching, that not only tackle childhood obesity but also enable parents to affect a lasting influence on their child’s health.

The study concluded that P-SES and KAP are more severe among families with boys compared to families with girls in China. The adverse impacts on BMI z-score and BFP appear significant among families with male preschool children. Several potential outcomes could arise from this research. Families with male children in China exhibit more severe Knowledge, attitudes, and practices linked to health and nutrition compared to families with female children, primarily because of the cultural desire for sons. This bias frequently results in increased expectations and resources allocated to boys, creating enormous strain on families to meet these demands, potentially resulting in negative socioeconomic consequences⁴⁰. Moreover, this cultural bias could impact the KAP related to health and nutrition in these households, possibly leading to a lack of emphasis on balanced food habits to prioritise the boy’s success⁴¹. This biased focus could unintentionally impact the BMI z-score and BFP of preschool children in families with boys more significantly. Boys may experience over-nutrition or under-nutrition due to misguided habits or high expectations, thus harming their physical health and development compared to girls.

Theoretical implications

Studying the impact of KAP in conjunction with parents’ socioeconomic status on preschool children’s BMI z-score and BFP has important theoretical implications for comprehending early childhood obesity. This approach emphasises the significance of a comprehensive framework for studying obesity, indicating that it is not just a consequence of personal behavioural decisions but is impacted by wider socioeconomic factors. The correlation between KAP and obesity indicators like BMI z-score and BFP in preschoolers implies that early treatments focusing on improving understanding and beliefs about nutrition and physical activity could be crucial. It questions current health behaviour models by highlighting the importance of early education and the socio-environmental setting in influencing children’s health results.

Secondly, socioeconomic factors highlight health disparity and how it appears early in life. The research aims to establish a correlation between P-SES and BMI z-score and BFP in preschoolers to demonstrate that health inequalities are socially influenced and manifest early in childhood. This discovery suggests a need to reassess theories about social determinants of health, emphasising a detailed understanding that considers age-specific vulnerabilities and the cumulative impact of socioeconomic disadvantage on children’s health. It emphasises the need for specific health policies and programs that target the underlying causes of these inequities, such as access to nutritious foods, safe spaces for physical exercise, and health education.

Furthermore, studying the influence of Knowledge, attitudes, practices, and socioeconomic factors on BMI z-score and BFP supports the developmental origins of health and disease (DOHaD) hypothesis. This viewpoint suggests that situations experienced in early life, such as during prenatal and early childhood, might have lasting impacts on health outcomes. Early exposure to inadequate diet and lack of physical activity, influenced by socioeconomic variables, can increase the likelihood of persons developing obesity and related health problems in the future. It emphasises the need to implement early preventative actions and advocates for a comprehensive strategy to promote health and prevent diseases throughout one’s life.

Finally, examining preschool children provides a distinct perspective for analysing the relationship between behaviour, environment, and physiology. By including BMI and BFP as outcome measures, the theoretical discussion on obesity is enhanced by recognising the intricacies of evaluating and understanding body composition in early childhood. It questions current paradigms that heavily depend on BMI z-score and promotes a more thorough approach considering body fat distribution and its health effects. This viewpoint promotes a transition to comprehensive health evaluation techniques in early childhood, acknowledging the constraints of BMI z-score as the only obesity indicator and supporting alternatives that better represent body composition and metabolic health.

Practical implications

This study presents numerous practical implications for health officials, educators, and community leaders aiming to tackle juvenile obesity via targeted interventions, especially within various socioeconomic circumstances. The study delineates specific strategies for implementation at the family, community, and institutional levels to alleviate the effects of P-SES on childhood obesity by enhancing P-KAP.

• Governments and public health agencies should develop and execute nutrition education initiatives specifically aimed at families of low socioeconomic status, emphasising cost-effective, nutritious meal planning and portion management.

• Despite budgetary limitations, community health workers could offer home-based or online educational sessions to instruct on practical, culturally relevant methods for integrating nutrient-dense foods into daily meals.

• Policymakers should implement subsidised nutritious meal initiatives to guarantee that economically disadvantaged families can get healthful food.

• Expanding food voucher systems or healthy school lunch programs can ensure that children from underprivileged homes obtain adequate nutrition at school and minimise their dependency on inexpensive, processed meals at home.

• Urban and rural community leaders should collaborate to establish secure and accessible environments for children’s physical activities, including public playgrounds, complementary sports programs, and community exercise events. Schools should incorporate organised physical education curricula that promote daily movement, particularly for children from low socioeconomic status backgrounds, who may have limited access to extracurricular sports opportunities.

• Paediatricians and family healthcare professionals should incorporate parental counselling sessions into standard child checkups to assist parents in making informed health decisions about their child’s nutrition and activity levels.

• Educational institutions and healthcare facilities might monitor trends in BMI and BFP over time and offer tailored interventions for at-risk youth.

• Extensive public health awareness efforts must be executed via social media, television, and community workshops to inform parents from diverse socioeconomic backgrounds about the risks of childhood obesity and the significance of healthy lifestyle practices.

Limitations and future research directions

A primary limitation of this study is its regional emphasis on Urumqi City, which may constrain the generalizability of the results to other regions of China. Considering the substantial socioeconomic disparities between urban and rural regions, the correlations among P-SES, P-KAP, and childhood obesity outcomes may vary across areas with differing economic development, healthcare access, and dietary habits. Urban regions such as Urumqi may possess enhanced access to healthcare services, nutritional education, and organised physical activities. In contrast, rural areas frequently encounter restricted healthcare access, distinct dietary practices, and a scarcity of recreational facilities for children. Future research should broaden the investigation to encompass various urban and rural locales throughout China to obtain a more extensive and representative dataset. Furthermore, comparative analyses between rural and urban populations may yield profound insights into how regional disparities influence juvenile obesity risks and guide more focused public health interventions.

Self-reported measures are inherently subjective. Social desirability bias is the inclination for individuals to present themselves favourably, whether they are aware of it or not. This might result in individuals overestimating or underestimating their psychological well-being or emotional intelligence. Moreover, self-reported statistics depend on the individual’s capacity to remember and precisely document feelings, activities, or experiences. This recall may be imperfect, suggesting potential inaccuracies in the material. Utilise qualitative methodologies, such as focus groups and discussions. Compared to a questionnaire, some methods can offer deeper insights into students’ experiences and perspectives.

Cross-sectional studies collect data at a particular moment in time. Without longitudinal data, cross-sectional research cannot demonstrate the progression of factors or the possible influence of specific changes in social support on music performance anxiety over a student’s time in university. It is advisable to employ a longitudinal study design to address this limitation. Longitudinal studies enable researchers to monitor the development and changes in social support, self-efficacy, and emotionality among university students over time. These inquiries additionally examine the connection between these changes and fluctuations in music performance anxiety.

This study presents bootstrapped confidence intervals to demonstrate the statistical importance of their findings; nonetheless, there is insufficient discourse regarding the practical implications of the reported effect sizes. Effect sizes are essential for assessing the practical significance of the examined relationships, as statistically significant results may possess negligible practical implications if the effect sizes are insufficiently large. If an enhancement in parental awareness results in only a slight decrease in a child’s BMI z-score, the intervention may lack significance at the population level. A comprehensive examination of the clinical or behavioural significance of the reported effects would augment the study’s relevance for policymakers and practitioners. Offering context, such as comparing the effect sizes to those found in similar interventions or real-world programs, would elucidate whether the identified relationships result in significant advancements in childhood obesity prevention initiatives.

Another limitation of the study is that we did not include families with more than one child between 3 and 6 years of age. This methodology was adopted to avoid bias by clustering effects and to guarantee that each case in the study was subjected to a proper statistical independence condition. Children in the same household may be more closely related than the average of any two individuals, and they are more likely to share the same environment. Comparing to parents with their singleton child (or similar parents who are with the same birth rank of their second or higher order child), those parents who have several children (other than their newborn) may have different information, beliefs, or behaviour(s) because of cumulative child raising experience or resource distribution. It may be pertinent in future research to include these families as a structured random factor, with statistical controls (e.g., multilevel modelling, nested design, etc.) to see the effects of family size on the child’s health.

Author contributions

H.H collected the study data. H.S.J & H.H wrote the article. H.S.J performed statistical analyses. H.S.J & H.H read the article and made the necessary checks for its correction. Then all of them approved the article.

Funding

This study has been done under the grant title: Study on the effect of family environment and children’s lifestyle on childhood obesity (Grant Number: RCZK202357).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

The survey was conducted with the approval of the University of Malaya Research Ethics Committee. The relevant guidelines and regulations were used to perform the research methods. Participants of the study were verbally informed about the purpose, objectives, and their right to participate, decline, or withdraw their participation in the research activities. Respondents have been notified that the information given was private and confidential and will only be used for academic purposes. Written informed consent was obtained from all respondents. Both participants and their legal guardians provided written informed consent.

Consent for publication

Not applicable.