| Objective:(1)We performed a Meta-analysis to explore the reproducibility and validity of FFQs and factors related to reproducibility and validity of FFQs.(2)In the Northeast Cohort Study of China(NEC-biobank),we developed a new food frequency questionnaire(FFQ)to estimate nutrient and food intake of participants in the Northeast of China.The aim of the study was to assess the reproducibility and validity of the FFQ.to obtain accurate and effective dietary information.(3)To assess the dietary intake quantity and quality of participants of the Northeast Cohort Study of China.(4)To explore dietary pattern among the participants of the Northeast Cohort Study of China and assess the association between dietary pattern and the prevalence of obesity,hypertension,diabetes mellitus and dyslipidemia,and to provide scientific basis and theoretical support for nutrition education intervention.Methods: A systematic literature review was performed using Pub Med and Web of Science databases.Pooled correlation coefficients(95% confidence interval)were calculated.Subgroup analyses were performed to investigate factors related to the reproducibility and validity of FFQs.We recruited participants from the Northeast Cohort Study of China.A total of 150 participants were included in our study.To assess the reproducibility of the FFQ,participants completed the first FFQ(FFQ1),and about three months later performed the second FFQ(FFQ2)in the same participant.Spearman correlation coefficients,intraclass correlation coefficients(ICC)and cross-classification analysis were used to assess the reproducibility of FFQ.Weighted diet record(WDR)and blood biomarkers have been regarded as the best reference method to assess the validity of FFQs.Spearman correlation coefficients,weighted kappa statistics and Bland-Altman analyses were used to assess the validity of FFQ.Moreover,the validation coefficients(VC)were calculated used the method of triads.Food and nutrient intakes by participants were measured using FFQ.Average food and nutrient intakes were compared with the recommended daily food intake according to the Chinese Residents’ Balanced Diet Pagoda(2016)and Chinese Residents’ Dietary Nutrient Reference Intakes(2013).In addition,Diet Balance Index(DBI),Chinese Healthy Diet Index(CHDI),Chinese Healthy Eating Index(CHEI)and Chinese Diet Quality Index(DQI)were used to evaluate the dietary quality of residents of Northeast Cohort Study of China.We used exploratory factor analysis(EFA)to establish dietary patterns among Northeast of China and assessed the association between dietary pattern and the prevalence of obesity,hypertension,diabetes mellitus and dyslipidemia.In addition,the simplified dietary pattern scores were calculated by summing the weighted standardized intake of food group with high factor loadings on the dietary pattern.To assess the assess the association between simplified dietary pattern scores and obesity,hypertension,diabetes mellitus and dyslipidemia.Results:1.A total of 123 studies comprising 20,542 participants were eligible for the Metaanalysis of reproducibility of FFQ.The pooled crude intraclass correlation coefficients(ICC)ranged from 0.499 to 0.803 and 0.499 to 0.723 for macronutrients and micronutrients,respectively.Energy-adjusted intraclass correlation coefficients ranged from 0.420 to0.803 and 0.507 to 0.712 for macronutrients and micronutrients,respectively.The pooled crude and energy-adjusted Spearman correlation coefficients ranged from 0.548 to 0.851 and 0.441 to 0.793,respectively,for macronutrients;and from 0.573 to 0.828 and 0.510 to0.744,respectively,for micronutrients.2.For the validity of FFQ,the literature search identified 130 articles that included 21,494 participants.After a Meta-analysis of the FFQs validity correlation coefficients of the included studies,this study showed that the range(median)of validity coefficients of the24 HRs as reference methods was 0.220-0.770(0.416),and the FRs was 0.173-0.735(0.373),which indicated FFQs were valid to assess the overall dietary intake in nutritional epidemiology studies.3.In the present study,a total of 131 participants were included in the reproducibility study.The crude Spearman correlation coefficients and ICCs for reproducibility ranged from 0.51 to 0.74(median value = 0.63)and from 0.17 to 0.67(median value = 0.61)for energy and nutrients,and ranged from 0.44 to 0.85(median values = 0.66)and from 0.26 to 0.86(median value = 0.60)for food groups,respectively.The average percentage of participants classified as extreme one-third was less than 10% for energy and all nutrients.The weighted kappa values ranged from 0.30 to 0.60.4.In the present study,a total of 125 participants were included in the validation study.The range of crude Spearman correlation coefficients between FFQ and WDR was ranged from 0.15 to 0.55(median value = 0.46)for energy and nutrients and ranged from 0.26 to0.70(median value = 0.48)for food groups.For most nutrients and food,extreme misclassification into the opposite third was below 15% and the weighted kappa values were above 0.20.By triangulation technique,we found that the validation coefficients of FFQ ranged from 0.14 to 0.70.And the values were above 0.3 for most nutrients,which indicated moderate or high levels.5.The average daily intake(gram)of grain/tuber,whole grains/beans,tuber,vegetables,fruits,egg,fish/shrimp,Livestock/poultry,Soybeans/nuts,milk and milk products were(682.32 ± 221.10),(30.44 ± 25.59),(50.77 ± 42.70),(208.28 ± 119.16),(213.64 ± 177.6),(35.13 ± 25.55),(19.36 ± 21.52),(60.12 ± 45.82),(75.12 ± 72.94),(93.12 ± 106.09).The percentage of participants met the recommended dietary intake was 8.18%,19.38%,28.91%,16.39%,24.19%,7.99%,8.48%,33.44%,11.59%,5.29%.6.The average daily intake of total vitamin A,thiamine,riboflavin,vitamin B6,vitamin B12,niacin,vitamin C and vitamin E were(455.04±378.98)μg RE,(0.66±0.38)mg,(0.95± 0.39)mg,(0.17 ± 0.25)mg,(217.29 ± 111.85)mg,(113.25 ± 67.99)mg and(5.07 ± 2.63)mg.The percentage of participants met the recommended vitamin intake was 61.79%,51.73%,74.57%,32.55%,7.19%,119.10%,113.25%,36.21%,54.32%7.The average daily mineral intake of residents in Northeast Cohort Study of China was(492.89 ± 241.24)mg for calcium,(991.67 ± 345.31)mg for phosphorus,(1918.96 ±814.67)mg for potassium,(1175.72 ± 610.52)mg for sodium,(296.84 ± 114.17)mg for magnesium,(296.84 ± 114.17)mg for iron,(20.51 ± 7.88)mg for zinc,(35.09 ± 16.94)mg for selenium,and(2.11 ± 1.42)mg for copper,(4.79 ± 1.54)mg for Mn were and(85.71 ±87.38)μg for iodine respectively.The recommended intake of these minerals in RNI was53.28%,137.95%,95.95%,82.05%,90.11%,264.79%,108.78%,58.48%,264.01%,106.41%,142.85%.8.We used DBI to evaluate the overall diet quality of residents and found that high bound score(HBS)was(23.38 ± 2.98),low bound score(LBS)was(14.40 ± 5.42),diet quality distance(DQD)was(37.78 ± 5.91).In addition,the moderate dietary imbalance,moderate excessive intake and moderate insufficient intake of residents were respectively 86.99%,55.81,and 29.74%.The mean score of CHEI in this study was(68.31 ± 9.14).The mean score of CHDI in this study was(70.2 ± 11.93),and percentage of individuals that CHDI reached 60 was 80.54%.The mean score of CHEI in this study was(68.31 ± 9.14).The DQI score ranged from-58.22 to 34.28,with an average score of(-21.19 ± 12.78),and94.13% of the residents with a total score less than 0.9.We extracted healthy dietary patterns,animal food dietary patterns,sauces and pickled products dietary patterns,beverages and desserts dietary patterns by exploratory factor analysis.With potential confounders adjusted,adults in the highest quartile of the healthy dietary patterns were at lower risk of hypertension(OR = 0.83,95%CI = 0.72-0.95)and hyperglycemia(OR = 0.39,95%CI = 0.32-0.49).With potential confounders adjusted,adults in the highest quartile of the animal food dietary patterns were at higher risk of general obesity(OR = 1.04,95%CI = 0.91-1.18),central obesity(OR = 1.24,95%CI = 1.13-1.36),hypertension(OR = 1.06,95%CI = 0.95-1.18),and dyslipidemia(OR = 1.05,95%CI = 0.95-1.15),while at lower risk of hyperglycemia(OR = 0.38,95%CI = 0.32-0.46).With potential confounders adjusted,adults in the lower quartile of the sauces and pickled products dietary patterns were at higher risk of general obesity(OR = 0.58,95%CI = 0.49-0.69),central obesity(OR = 0.58,95%CI = 0.51-0.66),hyperglycemia(OR = 0.65,95%CI= 0.52-0.800,and dyslipidemia(OR = 0.89,95%CI = 0.78-1.02).However,adults in the lower quartile of the sauces and pickled products dietary patterns were at higher risk of hypertension(OR = 1.29,95%CI = 1.16-1.44).With potential confounders adjusted,adults in the lower quartile of the beverages and desserts dietary patterns were at higher risk of general obesity(OR = 1.34,95%CI = 1.18-1.52),central obesity(OR = 1.59,95%CI = 1.45-1.74),hypertension(OR = 1.10,95%CI= 1.00-1.22),hyperglycemia(OR1.36,95%CI = 1.14-1.62),and dyslipidemia(OR = 1.46,95%CI = 1.33-1.60).Conclusion: 1.By Meta-analysis,we found that the reproducibility correlations of FFQ were greater than 0.5 for most nutrients.In addition,the validity correlation of FFQ on energy and most nutrient intake roughly ranged from 0.4 to 0.7 and from 0.3 to 0.6 when24-hour recalls and food records used as reference method,respectively.Within this range of reproducibility and validity coefficients,FFQ may be considered a reliable and validation tool to measure dietary intake.2.We found that the reproducibility correlation of FFQs was above 0.5 for most nutrients and food groups,and the correlation coefficients between FFQ and WDR were roughly between 0.3 and 0.7 for nutrients and groups.It indicated that FFQ developed for residents of Northeast China was reasonably reliable and valid for assessing most food and nutrient intakes.2.Generally speaking,the residents of Northeast Cohort Study of China were in moderately unbalanced state,and the major problems were both adequate food intake and excessive food intake.Refined grains met the recommended intake levels,but fruits and vegetables,fish and shrimps,dairy and soy products,and calcium intake are insufficient.3.Healthy dietary pattern was negatively correlated with hypertension and diabetes mellitus.The dietary patterns of beverages and desserts were positively correlated with obesity,abdominal obesity,hypertension,diabetes and dyslipidemia. |
PDF Full Text Request