Cluster analysis for food group consumption patterns in a national sample of Palestinian schoolchildren: Evidence from HBSC Survey 2013-2014

Abstract

Background: Promoting a healthy diet and lifestyle to reduce the national burden of nutrition-related problems among Palestinians requires an understanding of food consumption trends and patterns. Few studies have examined the food consumption patterns with the macro and micronutrient intakes and nutrition risk factors. The objective of this study was to study the food frequency and nutrient intake consumption patterns of Palestinian schoolchildren and their associations with the socioeconomic and risk factors. This is a national cross-sectional descriptive study conducted on Palestinian schoolchildren from the West Bank. The study examined the food consumption patterns of the macro and micronutrient intakes and nutrition risk factors among 1945 students aged 11-16 years. The data collected using the food frequency questionnaire and 24-hour recall that was administered by trained field workers. Food groups’ classification, nutrient intakes, body mass index (BMI) Z-scores, and socioeconomic di"erences were examined across the food groups’ patterns of consumption. We employed Z-score and K-Means cluster analysis to identify food consumption patterns and to examine factors associated with nutrient intakes. The food frequency results identified three food consumption clusters including the traditional, non-traditional, and mixed pattern. A total of 796 students (41%) were in traditional cluster, 458 (23.5%) in non-traditional cluster, and 691(35.5%) in mixed cluster. The nutrient intakes identified three clusters (High, Moderate, and Low consumption patterns) out of macronutrient, vitamins, and minerals categories. Most of the students located in the low consumption cluster for macronutrient, vitamins, and minerals clusters (66.9%, 67.7%, and 64 %) respectively. The traditional cluster was associated with healthy, non-obese, and physically active students and the non-traditional cluster was associated with unhealthy and obese students, but both shown significantly di"erent across the identified clusters. Imbalance in dietary intakes among schoolchildren reflects a lack of dietary diversity. High sugar, fats and oils, and beverages consumption, low consumption of grains, fruits, beans and legumes, and meat are noticed in Palestinian schoolchildren. The findings indicated the importance of considering the food groups' intake variations among Palestinian schoolchildren. As the segments relate to children’s health, nutrition diet programs should consider the high scores of non-traditional and mixed food consumption among school’s children.

Cluster Analysis and Classification Model of Nutritional Anemia Associated Risk Factors Among Palestinian Schoolchildren, 2014

“…This study utilizes data from the Health Behaviour in School-aged Children (HBSC) survey conducted by Al-Quds University and the Ministry of Health in the year 2013–2014, which aims to assess the nutritional, physical, and psychological health of Palestinian schoolchildren ( 34 , 39 , 40 ). The study sample included children aged 10–17 years who were enrolled in public and the United Nations Relief and Works Agency (UNRWA) for Palestinian refugee schools.…”

“…Furthermore, data mining and machine learning techniques for clustering and classifications have been improved as effective tools in the clustering and classification of risk factors associated with the health of children ( 29 , 30 ). The K-means, Decision Tree, Chi-square automatic interaction detection (CHAID), the K-Nearest Neighbors' algorithm (K-NN), and Classification and Regression Trees (CRT) clustering and classification models have been used by many research studies for predicting and identifying nutrition, lifestyle, and health diseases, such as obesity, diabetes, and anemia ( 31 – 34 ). In medical research, several ML clustering and classification, such as support vector machine, artificial neural network, and random forest techniques, are used for improving early detection and diagnosis of diseases ( 29 , 30 , 35 – 38 ).…”