FACTORS ASSOCIATED WITH UNDERWEIGHT, OVERWEIGHT AND OBESITY AMONG SCHOOL-AGE CHILDREN IN A HIGH-ALTITUDE REGION OF CHINA

Abstract

Childhood undernutrition and overnutrition are major public health problems among school-age children in high-altitude multi-ethnic regions of the People’s Republic of China.  In this study, we aimed to determine factors independently associated with underweight, overweight and obesity among school-age children in Qinghai Province, a high-altitude multi-ethnic region of western China, in order to inform efforts to prevent these conditions.  The students were chosen by multi-stage stratified cluster random sampling.  One primary school from each city or prefecture was selected and then one class per grade (grades 1-6) was selected from each study school and the parents of all the students in that class were invited to participate in the study.  Inclusion criteria for the students were enrollment in a selected class, being aged 7-12 years, being a resident of Qinghai Province for at least 3 years prior to the study, having parents able to communicate in the survey language and being willing to participate in the study.  Exclusion criteria for the students were not meeting inclusion criteria, being hospitalized during the previous month, having a chronic disease that could affect their weight, currently taking glucocorticoids and having a major weight change prior to the study.  The minimum number of students calculated to be needed for the study was 807.  Study subjects were a parent from each student invited to participate.  A parent of each student was asked to complete a self-administered questionnaire asking about demographic factors, socioeconomic status of the family, student early-life factors, student diet, student lifestyle (outdoor physical activity, electronic device use and sleep duration), parental knowledge and attitudes about nutrition and parental behaviors (parental perception of the student subject’s body-size category, parental use of food as a reward, parental restriction of snacks, parental physical exercise habits and joint parent-child exercise).  Parental knowledge about nutrition was assessed with 10 questions on the questionnaire.  A correct answer was given one pointand an incorrect answer was given 0 point.  The total possible score had a range of 0-10.  A higher score indicated better parental nutritional knowledge.  Parental attitudes about balanced nutrition were assessed using 11 questions answered using a 5-point Likert scale (each question had a score from 0 (disagreement with the statement) to 4 points (completely agreed with the statement) giving a total possible score of 0-0-    44 points, with a higher score indicating more favorable parental attitudes.  We measured each student’s height, weight and waist and hip circumferences.  Body mass index (BMI) was calculated as weight (in kg) / height (in m²).  Subjects were classified into four weight-status categories using age- and sex-specific cut-off points obtained from International Obesity Task Force criteria for underweight and Working Group on Obesity in China criteria for overweight and obesity.  The underweight, normal weight, overweight and obese criteria for boys aged 7, 8, 9, 10, 11 and 12 years were ≤13.9, 14.0-17.3, 17.4-19.1, ≥19.2; ≤14.0, 14.1-18.0, 18.1-20.2, ≥20.3; ≤14.1, 14.2-18.8, 18.9-21.3, ≥21.4; ≤14.4, 14.5-19.5, 19.6-22.4, ≥22.5; ≤14.9, 15.0-20.2, 20.3-23.5, ≥23.6; and ≤15.4, 15.5-20.9, 21.0-24.6, ≥24.7 kg/m², respectively; and for girls of the same ages were ≤13.4, 13.5-17.1, 17.2-18.8, ≥18.9; ≤13.6, 13.7-18.0, 18.1-19.8, ≥19.9; ≤13.8, 13.9-18.9, 19.0-20.9, ≥21.0; ≤14.0, 14.1-19.9, 20.0-22.0, ≥22.1; ≤14.3, 14.4-21.0, 21.1-23.2, ≥23.3; and ≤14.7, 14.8-21.8, 21.9-24.4, ≥24.5 kg/m², respectively. The study was conducted during September 2021-October 2022.  We used univariate and multivariate logistic regression analysis to identify factors significantly associated with underweight, overweight and obesity and the results are expressed as odds ratios (ORs) with 95% confidence intervals (CI).  A total of 1,561 student subjects were included in our study: 52.0% (n = 811) females.  The mean (±standard deviation (SD)) age of student subjects was 9.2 (±1.7) years.  279 student subjects (17.9%) were underweight, 1,065 (68.2%) had a normal weight, 122 (7.8%) were overweight and 95 (6.1%) were obese.  On multivariate analysis variables significantly associated with higher odds of the student subject being underweight were student subject male sex (aOR: 1.61; 95%CI: 1.22-2.12, p-value <0.001) and having a higher parental subject nutrition-related attitude score (for each
1-    point increase in the score there was an increase in the aOR of 1.04; 95%CI: 1.02-1.07, p-value = 0.002).  The factors significantly associated with lower odds of the student subject being underweight were: having been born by cesarean section (versus vaginal delivery; aOR: 0.58; 95%CI: 0.40-0.85, p-value = 0.005), having a parental subject perception of the student subject being in a higher BMI class than the student subject’s actual BMI class (for each higher BMI class than the actual student subject class the aOR decreased by 0.62; 95%CI: 0.49-0.78, p-value <0.001), more frequent parental physical exercise (for each one-level increase on the 4-level exercise scale the aOR decreased by 0.68; 95%CI: 0.54-0.87, p-value = 0.002) and more frequent parent-child joint exercise (for each one-level increase on the 4-level exercise scale the aOR decreased by 0.72; 95%CI: 0.52-0.98, p-value = 0.039).  The variables significantly associated with higher odds of the student subject being overweight (versus normal weight) were the student subject having been born by cesarean section (aOR: 1.80; 95%CI: 1.12-2.89, p-value = 0.015) and having a parental subject perception of the student subject being in a higher BMI category than the actual category (for each category higher than the actual category the aOR increased by 2.24; 95%CI: 1.72-2.92, p-value <0.001).  The variables significantly associated with lower odds of the student subject being overweight were older student subject age (for every 1-year increase in the student subject age the aOR decreased by: 0.87; 95%CI: 0.76-0.98, p-value = 0.028), having a higher parental subject nutrition-related attitude score (for every 1-point increase in the score the aOR decreased by: 0.95; 95%CI: 0.92-0.98, p-value = 0.002) and more frequent parental physical exercise (for each one-level increase on the 4-level exercise scale the aOR decreased by 0.64; 95%CI: 0.46-0.89, p-value = 0.009).  The factor significantly associated with higher odds of student subject obesity (versus normal weight), was having a parental subject perception of the student subject being in a higher BMI category than the actual BMI category (for each higher BMI category than the actual category the aOR increased by 1.83; 95%CI: 1.36-2.47, p-value <0.001).  The factors significantly associated with lower odds of student subject obesity were older student subject age (for every 1-year increase in age the aOR decreased by: 0.71; 95%CI: 0.62-0.83, p-value <0.001) and more frequent parental physical exercise (for each one-level increase on the 4-level exercise scale the aOR decreased by 0.45; 95%CI: 0.29-0.70, p-value <0.001).  In summary, factors significantly associated with greater odds of the student subject being underweight were student subject male sex and having a higher parental subject nutrition-related attitude score and with lower odds of being underweight were being born by cesarean section and having a parental subject perception of the student subject being in a higher BMI class than the actual BMI class, more frequent parental physical exercise, and more frequent parent-child joint exercise.  Factors significantly associated with higher odds of student subject being overweight were having been born by cesarean section and a parental subject perception of the student subject being in a higher BMI class than the actual BMI class and with lower odds of being overweight were older student subject age, higher parental subject nutrition-related attitude score and more frequent parental physical exercise.  The factor significantly associated with higher odds of student subject obesity was having a parental subject perception of the student subject being in a higher BMI class than the actual BMI class and lower odds of obesity were older student subject age and more frequent parental physical exercise.  We conclude, there is a need for education of study students and their parents regarding exercise and whether the student’s weight was normal or not.  Further studies are needed to determine how to improve both student and parental attitudes about exercise and nutrition.