Study sample

From the study of the Shanghai Children’s Health, Education and Lifestyle Evaluation-Adolescents (SCHEDULE-A), a population-based cross-sectional survey investigating adolescent physical and mental health (11–18 years), the current study utilized a subsample of adolescents recruited from Shangrao region (December 2018), a relatively social-economic underdeveloped city located in the downstream Yangtze River in southeast China. The subsample included both overall and abdominal weight indicators, and also had a relatively high underweight rate that allowed us to examine the underweight-EF relationship. A multi-stage cluster random sampling approach (i.e., District-School-Class-Student) was used. Details were reported in Additional file 1: Method S. Briefly, four districts/counties were firstly sampled according to the per capita disposal income of the Chinese residents in 2016; next, in each sampled district/county, two junior and two senior high schools were selected at random (4 districts/counties × 4 schools); then one class from each grade of the selected schools was drawn randomly; finally, all students were invited to take part in the study. The protocol was approved by the Shanghai Children’s Medical Center Human Ethics Committee according to the Declaration of Helsinki (SCMCIRB-K2018103). All parents and their adolescent children provided written informed consents.

Through the above-mentioned sampling method, 2704 students were selected, and 2346 students (86.8% response rate) agreed to participate. Finally, a total of 1935 students were retained after data cleaning based on the following inclusion criteria: (1) age ranged from 11 to 18 years; (2) without any chronic physical and mental disorders; and (3) without missing and invalid data of the main variables, i.e., weight, negative emotions, and EF (Fig. 1).

Fig. 1
figure 1

Flowchart of the participants. BMI body mass index, BRIEF the Behavior Rating Inventory of Executive Function, DASS the Depression Anxiety and Stress Scales with 21 items, SCHEDULE-A study of the Shanghai Children’s Health, Education and Lifestyle Evaluation-Adolescents, WHtR waist-to-height ratio

Main variables

Weight spectrum

We obtained information on adolescents’ height, weight, and waist circumstance from each school, which were measured by staffs of the school infirmary. BMI was calculated as the weight divided by the squared height, and waist-to-height ratio (WHtR) was computed as the waist circumference divided by the height. BMI value was subsequently converted into z-score using the least mean square method according to the World Health Organization growth reference [22]. Then, we defined BMI z-score <  − 1,  − 1 ~ 1, ≥ 1, and ≥ 2 as the overall weight spectrum, i.e., underweight, normal, overweight, and obesity, and defined WHtR < 0.40, 0.40 ~ 0.46, ≥ 0.46, and ≥ 0.50 as the abdominal weight spectrum, including thinness, normal, overweight, and obesity, respectively [23, 24].

Negative emotions

We assessed three types of adolescent negative emotional status (i.e., depression, anxiety, and stress) by self-report, using the Chinese version of the Depression Anxiety and Stress Scales with 21 items (DASS-21). This version was validated in mainland China with internal consistency indices (Cronbach’s alpha) of 0.83, 0.80, and 0.82, and test–retest reliability of 0.39, 0.43, and 0.46 for the depression, anxiety, stress domains, respectively, which supported its potential clinical utility in Chinese population [25]. All items were rated on a four-point Likert-type scale from 0 “did not apply to me at all” to 3 “applied to me very much or most of the time”. Each domain includes seven items. Specifically, item 3, 5, 10, 13, 16, 17, and 21 belong to depression domain, item 2, 4, 7, 9, 15, 19, and 20 belong to anxiety domain, and item 1, 6, 8, 11, 12, 14, and 18 belong to stress domain. The summed score for each domain was multiplied by 2, and higher scores indicated greater negative emotional symptoms. Adolescent who scored moderate or above with the cutoffs of ≥ 14, ≥ 10, and ≥ 19 were classified as experiencing depression, anxiety, and stress issues, respectively.

Executive function

We measured adolescent EF in a natural setting by parent report, using the Chinese version of the Behavior Rating Inventory of Executive Function (BRIEF), which can serve as a screening tool for possible executive dysfunction in children and adolescents aged 5–18 years old [26]. The BRIEF has eight non-overlapping domains, and almost all showed a good internal consistency (0.74–0.96) and test–retest reliability (0.68–0.89) in the Chinese population [27]. Of the total 86 items, each one is rated on a three-point scale (i.e., never, sometimes and often), and parents were required to select the most suitable answer for each described behavior of their adolescents in the past 6 months. For a few individuals aged over 18 years (n = 137), we utilized the same version, as they lived together with their parents, and their everyday function therefore could be well-assessed by their parents [28]. We checked the raw data based on two validity indexes (i.e., negativity < 5 and inconsistency < 7) to reduce reporting bias according to the BRIEF manual.

Eight domains of the BRIEF can form two broader indices and one overall index. The broad Behavior Regulation Index (BRI) includes three domains, i.e., inhibit, emotional control, and shift, which are interpreted as the ability to regulate one’s own behavioral and emotional control, and to move flexibly from one action to another. The broad Metacognition Index (MI) incorporates five domains, i.e., working memory, plan, initiate, organize, and monitor, which are related to the ability to solve problem actively, and to initiate, organize, and monitor one’s own actions. All eight domains form the overall index, i.e., the Global Executive Composite (GEC). T-scores were computed based on sex- and age-specific norms, with the T-score > 60 and > 65 being defined as potential sub-clinical and clinical EF impairments or problems.


According to a recently suggested principle of confounder selection, we selected several covariates that may influence the association between weight spectrum and EF, such as social-demographic factors (i.e. adolescents’ age, sex, parental highest education, and family gross income) and individual lifestyle behaviors (i.e. screen exposure, nighttime sleep duration, and physical activity) [29]. Parents or other main caregivers reported the socio-demographic information, including parents’ educational attainment, gross household income, and adolescents’ sex, birth date, and some chronic physical and mental health status. Screen exposure was measured by two widely used questions in children and adolescents [30], that is in the last month, on average, the total time he/she spent per day on (1) sitting and watching television or videos, and (2) playing games using device such as cellphone, iPad, PlayStation, etc. Each response was dichotomized, with ≥ 2 h/day indicating excessive screen time. Adolescents were also asked “At what time do you usually go to bed and get up on weekdays and weekends, respectively?” Averaged night sleep duration (i.e., time in bed in current study) was calculated by (5*weekdays + 2*weekends)/7, then was classified as shorter duration by cutoffs of < 9, < 8, and < 7 h for adolescents aged 12–13, 14–17, and ≥ 18 years, respectively [31]. We used a short Chinese version of the International Physical Activity Questionnaire to measure adolescent physical activity intensity, and then categorized it into low, moderate and high level [32].

Statistical analysis

Participants’ characteristics were described with mean (SD) and frequency (%), and the social-demographic differences between analyzed and excluded sample were assessed by t-test and chi-squared test for continuous and categorical variables, respectively.

In considering of a non-linear association between weight and EF, we fitted a linear regression model with BMI z-score and WHtR included as both a linear and quadratic term for each EF domain. Because almost all BMI z-score quadratic terms were not statistically significant (Additional file 1: Fig. S1), but WHtR quadratic terms were (Additional file 1: Fig. S2), two overall weight categories (i.e., normal and overweight), and three abdominal weight categories (i.e., thinness, normal, and overweight) were used in the subsequent analyses (because of a low obesity prevalence, these adolescents therefore were included in the overweight category, Table 1).

Table 1 Participant characteristics

We used multivariable logistic regression with standard error type of clustered robust to determine the association of weight spectrum and negative emotion with each sub-clinical EF problem (because of a low prevalence of clinical EF issues, adolescents with clinical EF issues, therefore, were included in the sub-clinical category, Table 1). The potential moderating effect of each negative emotion on the weight-EF link was tested by adding an interactive term (e.g., depression × overweight) in the model. Should the interaction term reached statistical significance, simple effect analysis was subsequently performed. We further conducted a post-hoc analysis of the associations between negative emotions and weight spectrum using multinomial logistic analysis. Finally, we performed a sensitivity analysis by limiting adolescents with WHtR values within 0.2–0.7 (removing possible outliers, Additional file 1: Fig. S2), and by omitting individuals with age over 18 years (the BRIEF was developed for children and adolescents aged 5–18 years).

All analyses were performed in Stata 15.0, and P values < 0.05 were considered statistical significance.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.


This article is autogenerated using RSS feeds and has not been created or edited by OA JF.

Click here for Source link (