Prevalence and incidence of MAFLD and associated anthropometric parameters among prepubertal children of the Shanghai Birth Cohort

Metabolic dysfunction-associated fatty liver disease (MAFLD) is the most common chronic liver disease in adolescent and adult population. However, the epidemiologic data of MAFLD in prepubertal children remain limited. This study aimed to investigate the prevalence and incidence of MAFLD and assess the role of anthropometric parameters in identifying and predicting MAFLD in this population. Children from the Shanghai Birth Cohort Study who underwent an 8-year follow-up with anthropometric measurements and transient elastography FibroScan-502 examination (M probe, Echosens, Paris, France) were enrolled. Some of them also completed a 5-year follow-up. Diagnosis of fatty liver disease (FLD) was based on the controlled attenuation parameter (CAP) value exceeding 248 dB/m, and MAFLD was defined as FLD combined with obesity or central obesity. Receiver operating characteristic (ROC) curve analysis was conducted to evaluate the diagnostic accuracy of anthropometric parameters for MAFLD. A total of 848 children (431 boys) from the Shanghai Birth Cohort Study were followed up for 8 years, and among them, 385 children (189 boys) also participated in the 5-year follow-up. The prevalence of FLD and MAFLD at 5 years old was 3.90% and 0.52%, respectively, while at 8 years old, the prevalence rates increased to 5.07% for FLD and 3.42% for MAFLD. The 8-year-old children with MAFLD exhibited significantly higher weight, body mass index (BMI), chest circumference, waist circumference, hip circumference, waist-to-height ratio, waist-to-hip ratio, and liver stiffness measurement compared to those without MAFLD (all p < 0.05). The incidence rates of FLD and MAFLD at 8 years old, considering the 5-year follow-up data, were 3.78% (14/370) and 3.13% (12/383), respectively. Obese or centrally obese children at 5 years old had a higher incidence of FLD and MAFLD at the 8-year follow-up. Waist circumference and BMI showed significant associations with the presence and incidence of MAFLD, respectively, with the largest AUC values in ROC curve analysis. In addition, chest circumference was significantly associated with MAFLD in obese children. This study provides insights into the incidence and prevalence of MAFLD in prepubertal children. It underscores the importance of anthropometric parameters in identifying and predicting MAFLD in this population. Further research encompassing a broader age range and incorporating these indicators and additional metabolic markers is necessary to enhance the understanding and management of MAFLD in children.


Introduction
The incidence and prevalence of overweight and obesity in children are increasing globally due to sedentary lifestyles and high-calorie diets, which pose significant public health concerns [1,2].Previous studies found that childhood obesity was associated with adult obesity and may even lead to severe complications in the future [3].Obesity and central obesity are both linked to higher risks of cardiometabolic health problems, such as metabolic syndrome, impaired insulin metabolism, and hypertension [4].Meanwhile, the prevalence of fatty liver disease (FLD) in children has increased significantly over the past few decades, due to the rising rates of childhood overweight and obesity.Non-alcoholic fatty liver disease (NAFLD) has become a prevalent cause of chronic liver diseases in children, affecting 10% of the general pediatric population and up to 40-70% of obese pediatric patients [5][6][7][8].NAFLD in children increases the risk of developing type 2 diabetes (T2DM) and cardiovascular disease in adulthood and can lead to liver-related death and significantly shorter survival [9][10][11].
Liver biopsy is the gold standard for diagnosis and staging of fatty liver in children, but it is invasive, costly, and not suitable for screening or follow-up assessments.Therefore, alternative non-invasive methods to liver biopsy have been increasingly studied.The device (FibroScan-502, Echosens, Paris, France), developed using the transient elastography (TE) technique, can simultaneously obtain controlled attenuation parameter (CAP) and liver stiffness measurement (LSM) in a rapid, non-invasive, reproducible, and painless manner [12].Numerous studies have shown that TE is a fast and reliable method for diagnosing liver fibrosis and NAFLD in adults, with several studies also reporting its utility in the pediatric population [13][14][15][16][17][18][19].Recently, the prevalence of FLD with CAP ≥ 248 dB/m in adolescent (12-18 years) in the USA was estimated using the Health and Nutrition Examination Survey 2017-18 database [18].
The term NAFLD is not appropriate for children, as alcohol consumption is typically not a contributing factor, and inherited metabolic disorders can present with similar symptoms or coexist with a diagnosis of NAFLD [20].In 2020, an international expert consensus panel suggested the term metabolic dysfunction-associated fatty liver disease (MAFLD) to replace NAFLD and introduced a simplified set of positive criteria for diagnosis [2,21].This redefinition has been validated by numerous studies, and in 2021, the criteria were adapted for pediatric practice [22,23].However, the specific features of MAFLD in young prepubertal children remain largely unknown.
In this article, the aim of our study is to utilize the TE technique to obtain CAP values and investigate the incidence and prevalence of MAFLD in young prepubertal children during their 8-year follow-up visit.We also aim to evaluate the role of commonly used anthropometric parameters in identifying and predicting MAFLD to aid in its diagnosis within this population.

Study population
The study population included 8-year-old children (94-98 months) from the Shanghai Birth Cohort [24], who completed follow-up of medical examinations and FibroScan measurement.In addition, a subset of participants also completed medical examinations and FibroScan measurements at the 5-year follow-up.Participants were excluded from the study if they met any of the following criteria: (1) lost to follow-up; (2) had missing anthropometric data or lacked available FibroScan data; (3) had any type of liver disease; (4) had additional liver diseases or were taking therapeutic drugs that could impact liver fat and function tests; or (5) experienced measurement failure of FibroScan-502 with M probe.The Ethics Committees of Xinhua Hospital affiliated to Shanghai Jiao Tong University School of Medicine approved the study, and informed consent was obtained from the parents of all participating children who signed written documents.

Anthropometric parameters assessment
The participating children underwent the medical examinations at the health examination center in Xinhua Hospital, where anthropometric measurements were obtained.Height was measured to the nearest 0.1 cm using stadiometers (Seca 416 Infantmeter, United States), and body weight was measured to the nearest 0.1 kg using digital scales (Detector 6745 Baby Scale, United States).Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters (kg/m 2 ).Chest circumference was measured by wrapping a measuring tape around the fullest part of the participants' chest, with the children standing straight and their arms relaxed.Waist circumference was measured by wrapping a measuring tape around the waist, above the participants' belly button, while they stood straight and identified their natural waistline.Hip circumference was measured by wrapping a measuring tape around the widest part of the participants' hips, while they stood straight with feet slightly apart.The measuring tape was snug but not tight, and it was held parallel to the floor.The measurements on the tape were read to the nearest 0.1 cm without compressing the skin.Waist-height ratio (WHtR) and waist-hip ratio (WHR) were calculated as waist circumference (cm)/height (cm) and waist circumference (cm)/hip circumference (cm), respectively.

Evaluation of liver steatosis and fibrosis using TE
After fasting for at least 6 h, all participants underwent FibroScan-502 with M probe (3.5 MHz) examination (Echosens, Paris, France) to estimate liver stiffness measurement (LSM) in kilopascals (kPa) and liver steatosis in decibels per meter (dB/m).CAP in dB/m and LSM in kPa were obtained simultaneously during each examination.The TE examination was considered successful when 10 valid measurements were obtained, with a success rate of at least 60%, and the interquartile range (IQR) was less than 30% of the median LSM value.Participants with unsuccessful examinations were excluded from the analyses.

Work definitions
Obesity was defined using age-and sex-specific cutoff points from the International Obesity Task Force, with BMI ≥ 95th percentile [25].Central obesity was defined using age-and sex-specific cutoff points for waist circumference ≥ 90th percentile [26].According to the definitions based on relevant references, a CAP value greater than 248 dB/m was considered for diagnosing FLD [18].For the diagnosis of MAFLD, in addition to meeting the diagnostic criteria for FLD, the presence of obesity or central obesity was also required [22].

Statistical analysis
Continuous variables were presented as mean ± SD for normal distribution and median ± IQR for skewed distribution.The t-test and χ 2 test were used to compare intergroup differences between different genders, weight groups, waist circumference groups, and FLD/ MAFLD groups for continuous and categorical variables, respectively.A univariate logistic regression model was utilized to identify potential anthropometric parameters associated with the prevalence or incidence of MAFLD, and a multivariate logistic regression model was used to identify the optimal index among all parameters that showed significance (p < 0.05) in the univariate analysis.To assess the performance of anthropometric parameters in identifying or predicting MAFLD, receiver operator characteristic (ROC) analysis was conducted.The area under the ROC curve (AUC), sensitivity, and specificity were estimated and compared.Statistical significance was considered at twosided p values < 0.05.The statistical analysis was performed using SAS Version 9.4.

Discussion
This is to our knowledge the first study to investigate the incidence and prevalence of MAFLD in prepubertal children using transient elastography.Our findings revealed that MAFLD was a common subtype within the spectrum of childhood FLD.As children advanced in age, there was a significant increase in the prevalence of both FLD and MAFLD, accompanied by a notable rise in the proportion of MAFLD within the FLD population.Moreover, the results indicated that certain metabolic dysfunctions, such as obesity and central obesity, may already be present before the onset of MAFLD in children.Importantly, we identified significant associations between anthropometric parameters and the occurrence and prevalence of MAFLD.In recent years, there has been a significant increase in the prevalence of childhood obesity, which is associated with various health risks, including an increased risk of T2DM, cardiovascular disease, and certain cancers [5].A study involving 626 early-stage adolescents reported an obesity prevalence of 16.3% [27].Central obesity, characterized by excess abdominal fat deposition, has also gained increasing attention due to its association with metabolic disturbances [28].In our study, we observed a prevalence of obesity and central obesity among 5-year-old children at 5.71% and 16.36%, respectively.During the 8-year follow-up, these rates significantly increased to 13.3% and 28.0%, respectively.While both boys and girls can be affected by  obesity, emerging evidences suggest sex-specific differences in the prevalence and causes of these conditions.Consistent with previous research [29], we found a significantly higher prevalence of obesity among boys (8.47%, 17.63%) compared to girls (3.06%, 8.87%) during both the 5-year and 8-year follow-ups.The underlying factors contributing to the gender disparity in obesity and central obesity prevalence are multifactorial and complex.Lifestyle differences between genders [30] may play a role, as studies have found that boys tend to have unhealthy eating habits and consume sweets more frequently than girls [31,32].It is worth noting that the increasing prevalence of obesity and central obesity underscores the need for comprehensive public health measures and targeted interventions to tackle this growing health issue among children.These findings also suggest the potential necessity of gender-specific guidelines for child nutrition.
The prevalence of FLD in children and adolescents has been increasing these years worldwide, primarily due to the obesity epidemic.Children with FLD are at risk of developing liver fibrosis, cirrhosis, and liver failure [33].Therefore, early detection, staging, and treatment of FLD in children are crucial in preventing disease progression.However, considering the unique characteristics of FLD in children, the term NAFLD may not be entirely appropriate.Pediatricians face challenges in diagnosing and managing these conditions, including the lack of clear diagnostic criteria, negative connotations associated with the current definition, and uncertainty regarding co-existing liver diseases [22,34].Furthermore, primary care physicians often overlook the prevalence of excess liver fat and associated risk factors in young children, leading to adverse long-term outcomes [35].To address these issues, the international panel of experts has proposed and adopted the MAFLD criteria for pediatric practice [22].It is crucial to validate the MAFLD diagnostic criteria across different age groups in children.In addition to liver biopsy, noninvasive methods such as serum alanine aminotransferase (ALT) levels or liver ultrasound (US) have been used in many pediatric studies to diagnose FLD [36,37].In previous studies, CAP measured by FibroScan has been found to be applicable in preschool children and useful in quantifying hepatic steatosis in pediatric patients [38,39].However, there is currently no consensus on the range threshold of CAP values for diagnosing FLD in children.In our study, we used a threshold of 248 dB/m for diagnosing FLD [18].In addition to meeting the criteria for FLD, the diagnosis of MAFLD in our pediatric cohort also required the presence of obesity or central obesity [22].Previous studies have reported a high prevalence of MAFLD in children and adolescents with obesity, estimated at 36.1% [40], and this prevalence increases with age and metabolic dysfunction category.In our study, the follow-up at 5 years old revealed a prevalence of FLD and MAFLD among children of 3.90% and 0.52%, respectively.However, as the children advanced in age, the 8-year follow-up showed a significant increase in the prevalence of FLD and MAFLD, reaching 13.33% and 27.95%, respectively.The proportion of MAFLD within the FLD population also increased from 15.38% at 5-year follow-up to 67.44% at 8-year follow-up, becoming the predominant subtype of FLD in children.At the 5-year follow-up, there was no significant difference in the prevalence of FLD and MAFLD between obese children (4.55%, 4.55%) and non-obese children (3.86%, 0.28%).However, at the 8-year follow-up, the results indicated a significantly higher prevalence of FLD and MAFLD among obese children (15.93%, 15.93%) compared to nonobese children (3.40%, 1.50%).Furthermore, we found that among the 8-year-old children, after the 5-year follow-up, the incidence rates of FLD and MAFLD were 3.78% and 3.13%, respectively.Among the newly diagnosed cases of FLD and MAFLD, when we traced back to the 5-year follow-up, there was already a significant increase in BMI, waist circumference, and a noticeable rise in the proportion of obesity and central obesity.These results emphasize the importance of screening FLD and MAFLD in children, particularly those with obesity or central obesity.In addition, certain metabolic dysfunctions, such as obesity and central obesity, may already exist before the onset of MAFLD in children.Hence, it is crucial to conduct follow-up screenings and implement appropriate intervention strategies for these children.
Given the increasing prevalence of MAFLD in childhood and adolescence, early diagnosis is important for facilitating timely intervention, impeding disease progression, and mitigating the risk of long-term complications [41,42].However, non-invasively identifying children with MAFLD is indeed challenging, and BMI alone may have limited utility in distinguishing between MAFLD and FLD, especially considering the increasing prevalence of childhood obesity.To improve the accuracy of MAFLD diagnosis, researchers have explored additional indicators such as waist circumference, serum lipid and blood glucose levels, as well as markers like the pediatric metabolic index and serum adipokine levels [22,43,44].Anthropometric parameters are easily obtainable in routine health examinations.Therefore, if significant indicators related to the occurrence and prevalence of MAFLD in children can be identified among these, it would provide convenient and rapid preliminary guidance to pediatricians during routine health assessments.In our study, we found that children with MAFLD at both 5 and 8 years of age exhibited larger BMI and waist circumference.Waist circumference showed a significantly associated with MAFLD prevalence in the overall participants at 8 years old and had the largest AUC of 0.872.In the obese subgroup, chest circumference had a significant association with MAFLD and the largest AUC of 0.813.For predicting MAFLD, BMI was the only parameter significantly associated with incidence in the multivariate analysis, with the largest AUC of 0.917.These findings suggest that anthropometric parameters may serve as good monitoring indicators for identifying and predicting MAFLD in this population.Future follow-up of this cohort and larger studies are needed to determine the roles of specific anthropometric parameters for identifying and predicting MAFLD in children of different ages.
Our study also has several limitations that need to be acknowledged.First, we did not conduct liver biopsy to confirm the absence of liver steatosis in the participants or use US as a comparison with CAP values.Nevertheless, liver biopsy is not ethically feasible for healthy children, and US is not considered accurate for diagnosing FLD in the pediatric population.CAP has been validated and increasingly utilized in various studies related to adult FLD and MAFLD, including pediatric research.Its application in our study provides valuable insights into studying the prevalence and incidence of MAFLD in prepubertal children.Second, the narrow age distribution of our cohort may limit the generalizability of our findings to other age groups.However, it is important to note that our study involves ongoing follow-up assessments.Future studies with a wider age range are necessary to obtain a more comprehensive understanding of MAFLD across different age groups.Third, we did not have data on important markers such as ALT, aspartate aminotransferase (AST), triglyceride high-density lipoprotein cholesterol (HDL-c), and glucose levels to diagnose related metabolic disorders.The inclusion of these parameters could further aid in differentiating MAFLD within FLD and identifying underlying metabolic dysregulation in seemingly healthy children.Future research should aim to incorporate these markers into more comprehensive disease risk stratification or prediction models to facilitate better follow-up, management, and early intervention for pediatric MAFLD.
In conclusion, our study provides valuable insights into the incidence, prevalence, and characteristics of MAFLD in young prepubertal children.The prevalence of MAFLD increases with age, and children with MAFLD exhibit increased BMI and waist circumference even before disease onset.Anthropometric parameters show positive associations with the incidence and prevalence of MAFLD, suggesting their potential as predictors during routine health assessments.Incorporating these parameters into practice may aid in early detection and intervention of MAFLD in this population.Further research is needed to explore their roles in different age groups and include additional metabolic markers for enhanced understanding and management of MAFLD in children.

Fig. 2
Fig. 2 Receiver operating characteristic (ROC) curves of anthropometric parameters for identifying metabolic dysfunctionassociated fatty liver disease (MAFLD) in perpetual children.a ROC curves of anthropometric parameters for identifying MAFLD in

Fig. 3
Fig. 3 Receiver operating characteristic (ROC) curves of anthropometric parameters for predicting metabolic dysfunctionassociated fatty liver disease (MAFLD) in perpetual children.a ROC curves of anthropometric parameters measured at the 5-year follow-up for predicting MAFLD at 8 years of age; b ROC curves

Table 1
Characteristics of 8-year-old children with and without hepatic steatosis/MAFLD p values are for Chi-square or Kruskal-Wallis tests between different groups.Means ± SDs were calculated for variables of normal distribution BMI body mass index, CC chest circumference, WC waist circumference, HC hip circumference, WHtR waist-to-height ratio, WHR waist-to-hip ratio, LSM liver stiffness measurement, kPa kilopascal, CAP controlled attenuation parameter, dB/m decibels per meter Parameters Total (n = 848) Non-FLD (n = 805) FLD (n = 43) p Non-MAFLD (n = 819) MAFLD (n = 29) p

Table 2
Univariate and multivariate logistic regression analyses showing the anthropometric parameters associated with MAFLD among 8-year-old children

Table 3
Characteristics of 5-year-old children with and without hepatic steatosis/MAFLD p values are for Chi-square or Kruskal-Wallis tests between different groups.Means ± SDs were calculated for variables of normal distribution BMI body mass index, CC chest circumference, WC waist circumference, HC hip circumference, WHtR waist-to-height ratio, WHR waist-to-hip ratio, LSM liver stiffness measurement, kPa kilopascal, CAP controlled attenuation parameter, dB/m decibels per meter

Table 4
Comparison of baseline parameters in 5-year-olds between subjects with CAP < 248 dB/m with incident FLD/MAFLD vs. those who did not develop FLD/MAFLD in 8-year-old follow-up (mean ± SD, unless otherwise stated) p values are for Chi-square or Kruskal-Wallis tests between different groups.Means ± SDs were calculated for variables of normal distribution BMI body mass index, WHtR waist-to-height ratio, WHR waist-to-hip ratio, LSM liver stiffness measurement, CAP controlled attenuation parameter

Table 5
Univariate and multivariate logistic regression analyses showing the baseline anthropometric parameters in 5-year-old children with incident MAFLD during follow-up