鈣質與鐵質攝取之飲食頻率問卷發展-青少年信度與效度研究

Development of a Food Frequency Questionnaire to Estimate Dietary Calcium and Iron Intake:Reproducibility and Validity Study for Adolescent

朱禹禎(Yu-Chen Chu)

指導教授 ： 林薇

國立臺灣師範大學/教育學院/人類發展與家庭學系/碩士(2006年)

摘要

本研究以台灣地區國人之飲食習慣為考量，擬發展出一份適合評估國人鈣質和鐵質攝取情況之飲食頻率問卷，並先就青少年族群進行信度與效度檢測。研究對象為台灣公立高中和國中學生，有效樣本共計193名，分別有96位高中生與97位國中生，其中男女學生人數各71名及122名。飲食頻率問卷之食物項目共八大類34項，其中用以評估鈣質攝取量之食物項目為28項，評估鐵質攝取量則有23項；使用份量參考圖片作為輔助工具。評估飲食頻率問卷之效度是以兩次三日飲食記錄做為效標；信度研究則比較二次間隔一個月之飲食頻率問卷。研究結果以t test、one way ANOVA、Pearson correlation、交叉分類及Kappa一致性等統計方法分析問卷之信度與效度。在信度研究方面，兩次飲食頻率問卷所測得之鈣質與鐵質攝取量的相關性分別達0.45和0.49，交叉分類結果顯示鈣質及鐵質攝取量分別有41%及39%以上受試者被歸在相同等分位，再以Kappa一致性係數來看，鈣質與鐵質攝取量分別呈現普通(0.22)和稍微(0.20)一致性；綜合研究結果，本研究之鈣質與鐵質飲食頻率問卷信度結果尚可接受。效度分析方面，經去除每日飲食變異性調整後，第二次鈣質飲食頻率問卷與兩次飲食記錄間相關性為0.48與0.51 (p< .01 )；而第二次鐵質飲食頻率問卷與兩次飲食記錄校正後相關性為0.33與0.32 (p< .01 )。為平衡飲食記錄資料收集順序所造成之測量誤差，將兩次飲食記錄結果平均，檢驗平均攝取量與第二次飲食頻率問卷之相關性，結果發現鈣質與鐵質之pearson相關性分別為0.50與0.33，皆達顯著相關(p< .01)。再將飲食記錄與飲食頻率問卷所測得之鈣質與鐵質攝取量，分別由高而低分為四等分位進行交叉分類比較，3成左右受試者其鈣質與鐵質攝取量在兩種評估方法比較後能正確分類至相同等分位。整體而言，鈣質飲食頻率問卷效度表現尚可，而鐵質飲食頻率問卷之效度較不理想，但仍屬可接受之範圍。

關鍵字

鈣質 ； 鐵質 ； 飲食頻率問卷 ； 飲食記錄 ； 信度 ； 效度

並列摘要

The purpose of this study was based on diet behavior of people living in Taiwan to develop a food frequency questionnaire (FFQ) of calcium and iron intake, and adolescent was the first conducted on the reproducibility and validity testing. The effective samples of this study are public senior high school and high school student with a total of 193. There are 96 senior high school students and 97 high school students respectively while the number of male are 71 and the number of female are 122. In FFQ, there are 34 food items in 8 categories, 28 items for evaluation of calcium intake and 23 items for evaluation of Iron intake, while the portion size was measured with referenced photos. Three-day food records were used as a reference standard for FFQ validity, and reproducibility of FFQ was tested 1-months apart. T test 、one-way ANOVA 、Pearson correlation 、Cross-classification and Kappa were used to evaluate the FFQ. For a test-retest reliability of the FFQ, the correlation for calcium and iron was respectively 0.45 and 0.49, and results showed that the cross-classification of calcium and iron intake were 41% and 39% of subjects were classified in the same quartile, then for Kappa statistic, calcium and iron intake were showing fair (0.22) and slightly (0.20) consistency. With these results, the reliability of this study for calcium and iron intake FFQs is acceptable. The correlation between calcium intakes as estimated by the second FFQ with two food records were 0.48 and 0.51 (deattenuated Pearson’s r), and the correlation between iron intakes as estimated by the second FFQ with two food records were 0.32 and 0.33 (deattenuated Pearson’s r). To balance the measurement error from the order of collecting food records, the two food records was averaged to examine the correlation with FFQ2, and the pearson correlation of calcium and iron are 0.50 and 0.33 (p <.01). Then participants were classified into quartiles according to their calcium and iron intake as estimated by two food records and FFQ, 30% of participants of their iron and calcium intake in the two assessment methods were correctly classified into the same quartiles. Overall, the validity of calcium FFQ was acceptable, and validity of iron FFQ was less desirable, but still acceptable

並列關鍵字

calcium ； iron ； food frequency questionnaire ； food record ； reproducibility ； validity

參考文獻

七年台灣地區膳食營養狀況。中華民國營養學會雜誌，16，39-60。

、磷、鐵、鈉）之食物來源：1993~1996國民養營健康狀況變遷調查結

，中華民國營養學會雜誌，3，11。

楊國樞、文崇一、吳聰賢、李亦園編（1981）。社會及行為科學研究法。台

余雅萍（2003）。學童版中式飲食頻率問卷之發展研究。國立台灣師範大學