重庆地区人群半定量食物频数问卷的效度和信度研究

Validity and reproducibility study of a semi-quantitative food frequency questionnaire in Chongqing adults

目的 对重庆地区人群半定量食物频率问卷 (SQFFQ)进行效度和信度研究。方法 采用 3日食物称量法为“标准”方法 ,对称量与半定量食物频率问卷结果进行相关性分析 ,检验其效度 ;间隔 4周进行两次问卷调查 ,比较两次结果之间的相关性 ,检验问卷的信度。结果 效度研究中 ,3 5种营养素SQFFQ与称量结果以原始数值计算的相关系数从 0到 0 .79,平均值 0 .40 ,经对数转换、能量校正或调查日间及个体间差异消减校正后的平均相关系数分别为 0 .41、0 .2 9和 0 44 ;信度研究中全部 3 5种营养素均为正相关 ,以原始数值计算的相关系数为 0 .3 0～ 0 .85 ,平均值 0 .67,经对数转换、能量校正后的平均相关系数分别为 0 .69和 0 .5 4,效度、信度研究吻合度分别为 82 .9%和 10 0 %。结论 所制订的重庆地区人群SQFFQ具有较好的效度和信度 ,在研究重庆市相关人群的疾病与营养素摄入之间关系时 。

Objective To evaluate the relative validity and reproducibility of a semi quantitative food frequency questionnaire to be used in prospective study of diseases and dietary factors among middle aged people in Chongqing. Methods Correlation analysis between nutrients intake in SQFFQ and in a 3 day weighed food record(WFR) employed as reference method was used to assess the relative validity. Then interviewer administrated SQFFQ survey was completed twice at a 4 week interval. Data were compared for the assessment of the reproducibility. Results Crude pearson correlation coefficient of 35 nutrients ranged from 0 to 0.79 in validity evaluation, with average value of 0 40, 0.41, 0.29, 0.44 for the crude, log transformed, energy adjusted and de attenuated coefficient, respectively. Mean crude, log transformed and energy adjusted correlation coefficients for reproducibility were 0.67, 0.69 and 0 5. The nutrient coincidence was 48.6% and 100% for validity and reproducibility, respectively. Conclusion The good validity and reproducibility suggest proper applicability. The questionnaire designed by our department might be an ideal evaluation tool in the studies of nutrients and diseases.