Development and application of statistical methods to correct for nonresponse bias in the Japan National Health and Nutrition Survey

2017 Fiscal Year Final Research Report

• Project/Area Number: 15K08762
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Epidemiology and preventive medicine
• Research Institution: National Institutes of Biomedical Innovation, Health and Nutrition
• Principal Investigator: Ikeda Nayu 国立研究開発法人医薬基盤・健康・栄養研究所, 国立健康・栄養研究所 国際栄養情報センター, 研究員 (20573603)
• Co-Investigator(Kenkyū-buntansha): 西 信雄 国立研究開発法人医薬基盤・健康・栄養研究所, 国立健康・栄養研究所 国際栄養情報センター, センター長 (80243228)
• Project Period (FY): 2015-04-01 – 2018-03-31
• Keywords: 国民健康・栄養調査 / 国民生活基礎調査 / 非協力者バイアス / 統計手法 / 保健統計 / 非感染性疾患 / レコードリンケージ / シミュレーション

Outline of Final Research Achievements

The response rate of the Japan National Health and Nutrition Survey (NHNS) is 60-70%. We examined statistical methods to correct for nonresponse bias in health data obtained from the NHNS. Our results suggest that household participation status be associated with household characteristics and that individual participation rates continue to decrease until 2045. Moreover, nonresponse bias in smoking prevalence estimated from NHNS participants was limited across age groups and sexes. Multiple imputation for survey nonresponse obtained more stable trends in upper-body obesity prevalence by smoking status compared with estimates from original data, and differences in upper-body obesity prevalence attenuated through corrected waist circumference and body mass index in smokers. In conclusion, the current response rates should be at least maintained, and multiple imputation may be useful to correct for nonresponse bias in anthropometric estimates in the NHNS.

Free Research Field

国際保健統計