Establishment of a method for predicting and preventing the risk of disease occurrence based on multiple imputation data of missing values of long-term health checkup examinees.

2021 Fiscal Year Final Research Report

• Project/Area Number: 19K11746
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 59040:Nutrition science and health science-related
• Research Institution: University of Shizuoka
• Principal Investigator: Kuriki Kiyonori 静岡県立大学, 食品栄養科学部, 教授 (20543705)
• Project Period (FY): 2019-04-01 – 2022-03-31
• Keywords: 生活習慣病 / 罹患リスク / 予測モデル / 時系列データ / 欠損値 / 多重代入法

Outline of Final Research Achievements

As part of the Japan Multicenter Cohort (J-MICC), data were collected on approximately 6,400 individuals in the Sakuragaoka area, Shizuoka, Japan, over a period of about 10 years. The Multiple Imputation by Markov Chain Monte Carlo (MCMC) method was used to complement missing values and to examine methods to predict and prevent the risk of developing obesity, hypertension, dyslipidemia, diabetes, and metabolic syndrome diseases. Electrolytes in urine samples were measured to maintain and control adequate blood pressure levels. As a model for analysis using the MCMC method in this study, we found that green tea, human gut microbiota, and season had an effect on blood glucose levels in the data analysis of the J-MICC Sakura Diet Survey (Four Seasons Diet Survey).

Free Research Field

栄養疫学

Academic Significance and Societal Importance of the Research Achievements

本研究は、『継続受診者』と『非継続受診者』の人口学的特性、各健診項目、生活習慣要因、精神ストレスなどを比較することから、継続受診の重要性を示すだけでなく、メタボリックシンドロームや高血圧などの一次予防や血圧維持のための食事・保健指導の方法を確立するうえで重要な基礎資料となる。大規模な長期の時系列データで疾患の罹患を高精度に予測するモデルを確立するにあたり、従来では、(尿試料中電解質や) 便試料の腸内細菌叢は検討されてこなかった。高精度の予測モデルに基づき、これらのバイオマーカーを用いて、食事・保健指導で生活習慣病を一次予防することの重要性を示すことに寄与した。