Qualitative and Semi-Quantitative Testing and Analysis of Sugar Content in Several Sweetened Beverages

¹ Faculty of Industrial Technology, Gunadarma University, Depok 16424, West Java, Indonesia
² Faculty of Health Sciences and Pharmacy, Gunadarma University, Depok 16424, West Java, Indonesia
³ Faculty of Computer Science and Information Technology, Gunadarma University, Depok 16424, West Java, Indonesia

^* Corresponding author: This email address is being protected from spambots. You need JavaScript enabled to view it.

Abstract

The increasing popularity and availability of sweetened beverages has contributed to a higher risk of metabolic diseases. This study analyzed the sugar content of various sweetened beverages using sequential laboratory tests. Qualitative analysis employed glucose paper strips to detect glucose presence, while semi-quantitative analysis used a refractometer to estimate total sugar content based on a standard curve. A total of 53 samples was selected due to their accessibility in the market and frequently consumed by the community, classified into nine categories by subject matter experts, were tested. Qualitative results indicated that 33 samples contained glucose, whereas 20 samples did not. Semi-quantitative analysis revealed that estimated total sugar content often exceeded the values stated on packaging labels, particularly in health supplement drinks, ready-to-drink coffee, and ready-to-drink tea. Several samples lacked sugar content information on their packaging despite exhibiting high sugar levels. The highest estimated total sugar content was observed in the energy drink category, followed by ready-to-drink tea and fruit juice. These results underscore the necessity for increased consumer awareness regarding excessive sugar intake and highlight the importance of regulatory oversight of nutritional labeling. Further analysis utilized a data mining approach, specifically the K-Medoids Clustering algorithm. The dataset was represented in two dimensions: qualitative features (glucose test results, binary Yes/No) and quantitative features (estimated total sugar content). Silhouette Score evaluation determined that three clusters were optimal. The first cluster comprised all samples without glucose, while the remaining two clusters separated glucose-containing samples by high and low sugar content. These results demonstrate the potential of data mining techniques to enhance sugar content analysis and characterize sweetened beverage test data.