Data Checking and Standardization in a Weighed Food Dietary Record Survey

• Published in: Eiyōgaku zasshi Vol. 58; no. 2; pp. 67 - 76
• Main Authors: Imaeda, Nahomi, Tokizane, Masami, Sato, Nobuko, Fujiwara, Nakako, Kamae, Michiyo, Tokudome, Yuko, Nagaya, Teruo, Maki, Shinzo, Tsunekawa, Suzue, Koide, Yayoi, Tokudome, Shinkan, Miyai, Yoshimi
• Format: Journal Article
• Language: English
• Published: The Japanese Society of Nutrition and Dietetics 2000
• Subjects: data checking; food coding; model recipe; weighed dietary record survey
• ISSN: 0021-5147; 1883-7921
• DOI: 10.5264/eiyogakuzashi.58.67

We collected weighed dietary records (WDRs) in four periods of 7 consecutive days in each-season from 1996 to 1997 (28 days in total) from 107 Japanese dietitians, and compiled 2, 925-day WDRs and 149, 187 lines for foods and weighed data. Of the 1, 160 foods consumed, 414 could not be suitably coded because they were precooked or unlisted in the Japanese Standard Composition Table (JSCT). There were 1, 112 coding errors in 73, 266 lines reported in the two surveys conducted in the spring and summer. We found that the database which had been specifically developed to check foods that are often erroneously coded was beneficial for correcting the coding errors. Verifying the data from interviews about dietary records with a standardized manual was found to be prerequisite for accurate food coding and weighing. Either overestimation or underestimation of the nutritional intake could occur with JSCT-unlisted foods, and model recipes for commonly consumed dishes therefore appear necessary. In order to ensure valid WDR surveys, a standardized manual with guidelines for manual and computer-assisted data checking, coding, data processing, and preparing standard recipes is recommended.