<![CDATA[This study aims to explore the pattern of diabetes mellitus risk factors among patients at Torabelo Regional Hospital in Sigi using Multiple Correspondence Analysis (MCA). A total of 465 patients were analyzed based on eight categorical variables, including age, gender, blood glucose levels, and lipid profiles. MCA was applied to identify inter-category relationships and visualize them in a low-dimensional space. The results show that most diabetes patients were female, aged 46 years and above, and had high fasting glucose, low HDL, and high LDL levels. The analysis identified two main patterns: a group with a low-risk metabolic profile who were not diagnosed with diabetes, and a group with a combination of high-risk metabolic categories who were more likely to already have diabetes. A distinct subgroup with extremely high triglyceride levels was also identified, indicating a rare but significant metabolic pattern. The first two dimensions of the MCA explained more than 40% of the data variation, providing sufficient support for meaningful visual interpretation. These findings demonstrate that MCA is effective in simplifying complex categorical data and supports risk-based segmentation strategies for early intervention planning in primary healthcare services, particularly in regions with high diabetes prevalence.]]>
Copyrights © 2025
