北京市儿童青少年食物频率问卷的信度和效度研究

Reproducibility and validity of a food frequency questionnaire developed for children and adolescents in Beijing

目的对研制的北京市儿童青少年食物频率问卷的信度和效度进行评价,为儿童青少年营养状况评估和健康干预以及饮食习惯与慢性疾病关联性的研究提供一个简便、可信的研究工具。方法对北京市130名10-17岁儿童青少年进行4次24h饮食回顾调查和2次食物频率问卷调查,24-h饮食回顾作为＂标准＂方法与食物频率问卷比较来评价其效度,2次间隔6个月的食物频率问卷调查结果相比较来评价其信度。结果两次食物频率问卷调查结果间能量及营养素摄入量估计差异较小,除蛋白质、磷和锌的摄入量第二次低于第一次外（P〈0.05）,其他营养素摄入量间差异无统计学意义（P〉0.05）。两次食物频率问卷调查结果间所有能量及营养素摄入量均呈正相关（P〈0.01）,Pearson相关系数从0.315（钙）至0.521（锌）,平均为0.431,女性高于男性（0.49vs 0.32）。食物频率问卷调查的能量及营养素摄入量均高于24h饮食回顾（P〈0.01）,食物频率问卷高估明显。调整总能量摄入和个体内变异等影响后,食物频率问卷与24h饮食回顾调查结果的能量和所有营养素摄入量均呈正相关（P〈0.05）,Pearson相关系数从0.27（维生素A）至0.53（锌）,平均0.38。四分位分组显示：研究对象被良好区分（相同组或相邻组）,比例从66.2%（维生素E、钙）至79.2%（铁）,平均73.0%,研究对象被严重错分（相隔2组）,比例平均为6.2%。结论此食物频率问卷有较好的信度和效度,可作为今后儿童青少年膳食营养摄入状况调查和评价的适用工具。

Objective To evaluate the reproducibility and validity of a food frequency questionnaire（FFQ）among children and adolescents in Beijing. Methods A total of 130 school children and adolescents aged 10-17years old were enrolled in Beijing,China.Two FFQ surveys and four 24-hour recalls were collected over a 6-month period.Reproducibility was evaluated by comparing two FFQ surveys;validity was evaluated by comparing the 24-hour recalls against FFQ data. Results Except for protein,phosphorus and zinc,intakes of the other nutrients were not significantly different between two FFQs（P〈0.05）.Nutrient intakes between two FFQs showed positive correlation（P〈0.01）.Pearson correlations ranged from 0.32（for calcium）to 0.52（for zinc）;and better in girls than in boys（0.49 vs 0.32）.Regarding the FFQ＇s validity,nutrient intakes from FFQs were greater than 24-hour recalls（P〈0.01）.After adjusted for total energy and intra-individual variation,all nutrient intakes showed positive correlation（P〈0.05）,and the correlations became stronger.The adjusted correlations ranged from 0.27（for vitamin A）to 0.53（for zinc）,with a mean of 0.38.For quartiles of the intakes,rates of FFQs and 24-hour recalls agreement were between 66.2%（for vitamin E and calcium）and 79.2%（for iron）,and misclassification to an extreme quartile was low（6.2%）. Conclusions The reproducibility and validity of the FFQ are modest and acceptable.It is feasible to use FFQ to assess the target population＇s dietary intakes.