Health Checkup Data Analysis Focusing on Body Mass Index

Mizuki HIGUCHI  Kenichi SORACHI  Yutaka HATA  

IEICE TRANSACTIONS on Information and Systems   Vol.E100-D   No.8   pp.1634-1641
Publication Date: 2017/08/01
Publicized: 2017/05/19
Online ISSN: 1745-1361
DOI: 10.1587/transinf.2016LOP0009
Type of Manuscript: Special Section PAPER (Special Section on Multiple-Valued Logic and VLSI Computing)
Category: Soft Computing
Body Mass Index,  lifestyle-related diseases,  specific health checkup,  Wilcoxon Rank-Sum test,  

Full Text: PDF(1MB)>>
Buy this Article

This paper analyzes the relationship between the changes of Body Mass Index (BMI) and those of the other health checkup data in one year. We divide all data of the subjects into 13 groups by their BMI changes. We calculate these variations in each group and classify the variations into gender, age, and BMI. As the result by gender, men were more influenced by the changes of BMI than women at Hb-A1c, AC, GPT, GTP, and TG. As the result of classification by age, they were influenced by the changes of BMI at Hb-A1c, GPT, and DTP by age. As the result of classification by BMI, inspection values such as GOT, GPT, and GTP decreased according to the decrement of BMI. Next we show the result on gender-age, gender-BMI, and age-BMI clusters. Our results showed that subjects should reduce BMI values in order to improve lifestyle-related diseases. Several inspection values would be improved according to decrement of BMI. Conversely, it may be difficult for subjects with under 18 of BMI to manage them by BMI. We show a possibility that we could prevent the lifestyle disease by controlling BMI.