Meta-analysis of Non-Linear Dose-Response Curves and its Application

2017 Fiscal Year Final Research Report

• Project/Area Number: 15K00052
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Statistical science
• Research Institution: Miyazaki Prefectual Nursing University
• Principal Investigator: Nakao Hiroyuki 宮崎県立看護大学, 看護学部, 教授 (40336293)
• Co-Investigator(Kenkyū-buntansha): 高橋 邦彦 名古屋大学, 医学系研究科, 准教授 (50323259)
• Co-Investigator(Renkei-kenkyūsha): HATTORI Satoshi 大阪大学, 医学系研究科, 教授 (50425154)
• Project Period (FY): 2015-04-01 – 2018-03-31
• Keywords: メタアナリシス / 疫学 / 用量反応関係

Outline of Final Research Achievements

In epidemiological studies, it is often necessary to determine the relationship between exposure levels and the risk of disease. Exposure levels are often available in intervals. Recently we proposed a procedure for assessing nonlinear associations between exposure levels and the risk of disease from a summarized grouped data using a restricted cubic spline model with 3 knots. We applied this procedure to perform a meta-analysis for the association between alcohol intake and the log RR of total mortality based on summarized data that was used in Di Castelnuovo et al (2009). We would demonstrate the shape of the estimated curves by the cubic spline model, and assess the differences between our results and those of Di Castelnuovo et al, in application of our procedure to epidemiological studies.

Free Research Field

疫学・統計学