24시간 회상법과 식품섭취빈도조사법의 커피 섭취 비교: 도시기반 코호트 활용

• Published in: Korean journal of community nutrition Vol. 25; no. 1; pp. 48 - 60
• Main Authors: 김안나(An Na Kim), 윤지영(Jiyoung Youn), 조현정(Hyun Jeong Cho), 진 타이위에(Taiyue Jin), 신상아(Sangah Shin), 이정은(Jung Eun Lee)
• Format: Journal Article
• Language: Korean
• Published: 대한지역사회영양학회 2020
• Subjects: food frequency questionnaire; the Health Examinees study; comparison; 24-hour recalls; coffee
• ISSN: 1226-0983; 2287-1624; 2951-3146
• DOI: 10.5720/kjcn.2020.25.1.48

Objectives: Most cohort studies used food frequency questionnaires (FFQ) to evaluate coffee consumption as it assesses habitual dietary patterns, whereas some studies have used the 24-hour recalls (24HR) as it elicits in-depth description of foods and the amount eaten. The aim of this study was to compare FFQs and 24HR to assess the consumption of various types of coffee. Methods: We included 25,904 participants aged 40 years or older from the Health Examinees (HEXA) Study of the Korean Genome and Epidemiologic Study (KoGES). Each participant completed one FFQ and one-day (n=11,280) or two-day 24HR (n=14,624). We classified coffee types into: black coffee, coffee with sugar and cream, and coffee with sugar alone or cream alone. We compared the proportions of nondrinkers, black coffee, and coffee with sugar and cream through FFQ and 24HR. Results: Among those who completed one FFQ and one-day 24HR, 39.4% of "nondrinkers" on one-day 24HR reported that they did not drink coffee on their FFQs. Whereas among those who complete two-day 24HR, 71.2% of "nondrinkers" on two-day 24HR said that they did not drink coffee on their FFQs. Among those who completed one FFQ and oneday 24HR, 58.3% marked "black coffee" on one-day 24HR said that they drank black coffee on their FFQs. Among those who complete two-day 24HR, 58.8% marked "black coffee" on two-day 24HR said that they drank black coffee on their FFQs. The kappa coefficients and percent agreements were 0.4 and 59.6%, respectively, for the comparison of coffee intake between FFQ and one-day 24HR, and 0.6 and 72.8%, respectively, for the comparison of coffee intake between FFQ and two-day 24HR. Conclusions: We found discrepancies between FFQs and 24HR in the types of coffee consumed. Such limitations should be considered when using the 24HR data to examine the effect of coffee consumption on disease development.