Comparison of Relative Validity of Nutrient Intake in Japanese Adults - Article Example

Comparison of Relative Validity of Nutrient Intake in Japanese adults Describe briefly the main purpose of the study and how the objectives were accomplished. The main purpose of the study was to investigate the relative validity of food group intakes in Japanese adults (5). The study involved ninety-two Japanese women of between 31 and 69 years and ninety-two Japanese men of between 32 and 76 years (12). The objective of the study was accomplished by subjecting the participants to a comprehensive self-administered diet history questionnaire and a brief-type diet history questionnaire (BDHQ). The study adopted the semi-weighted dietary records as a reference method (8). 2. The appropriateness of the statistical approach The study used the SAS statistical software to carry-out the statistical analyses of the data (66). Further, all the analyses were performed on log-transformed values thereby accounting for any non-normality (77). The SAS statistical approach provided for the calculation of energy-adjusted values using regression and also used the Pearson correlation coefficients hence making the results more valid and reliable (78). The nutrient intakes deprived from the first DHQ and the first BDHQ was compared with the intake from the 4 day DRs. (80). The correlation coefficients were then compared using the Meng-Rosental-Rubin method to identify overlapping correlation coefficients (93). Every questionnaire was evaluated using a paired t-test with two sided values, with p˂0.05 being considered as the significant difference (88). Apart from the use of Pearson’s correlations, another method of statistically assessing validity is by using Spearman’s correlation coefficients. I would also use biomarkers as a reference method in some of the nutrients, particularly those with larger intra-subject variation. This method would be relatively economical and less time consuming. 3. The time sequence of events for data collection (48) The study design was appropriate in assessing the relative validity of the DHQ and BDHQ. First, through the DHQ, the statisticians were able to estimate the dietary intake, behavior and consumption frequency of the participants during the preceding period. The study design used the DHQ that was able to ask for the consumption frequency and the portion sizes of the 150 food and beverages (60). On the other hand, the BDHQ provided the researchers with the opportunity to compare the dietary intake of some selected foods, although not based on the portion size (58). In addition, the reasonable sample size for a validation study is between 200-300 participants with three days of dietary measurements. In this case, the number of participants was 182 and the dietary measurement days were 4 which fall within the reasonable sample size. 4. The two energy-adjusted methods are relatively similar in results. From the findings, in the case of the crude intake of the 42 nutrients and the residual method, the researchers observed that the differences were insignificant. It was observed that energy adjusted nutrients by residual methods produces better results for women while energy adjusted by the density method produces similar outcomes for both men and women. From the first method, the DR and BDHQ1 nutrients intake were at 67% and 43% for women respectively and 26% and 64% for men respectively (101). For the energy adjusted for nutrient density, there was no significant differences that was observed in 13, 16 for the DHQ, BDHQ. Conceivably, it means that the significant difference was in 39, 22, 30, 27 nutrients out of the 43 used in the study. 5. From table 1, out of 43 nutrients, the table below shows the number of nutients that were either underestimated or overestimated during energy adjustment by residual and energy adjustment by density between DR and DHQ in women and men. The number of nutrients underestimated or overestimated Energy adjustment by residual between Energy adjustment by density between Total no. of nutrients DR and DHQ DR and BDHQ DR and DHQ DR and BDHQ Women 43 14 25 21 34 Men 43 32 16 30 27 From the above table, it can be noted that it is not easy to make a conclusion on which energy adjustment method produced better results. However, it appears that energy adjustment by residuals produced better results, particularly while comparing DR and BDHR. 6. In general, the results in 4 and 5 shows that there were no significant statistical differences between the DHQ1 and the BDHQ1. In particular, the data shows an insignificant statistical correlation while comparing the DR with the DHQ and BDHQ in the majority of the nutrients. 7. From Table 2, the Spearman’s correlation coefficients for women for the DHQ1 correlations across crude, energy-adjustment by residual and by density correlations were 0.43 with a range of -0.09 to 0.77. However, for the BDHQ1 in women the correlation was 0.44 with a range of 0.14 to 0.82. There is an insignificant statistical variation in the 0.43 and 0.44 coefficients for the DHQ1 and BDHQ1 respectively. On the other hand, for the men, the correlations are 0.44 with a range of 0.08 to 0.48 and 0.48 with a range of 0.22 to 0.83 for the DHQ1 and BDHQ1 respectively. This correlation pattern shows a reasonable ranking ability. 8. The de-attenuated correlations are statistically significant since they show the correlation between the nutrient in-take while taking into consideration the effect of within-person variations in the data that were recorded. Accordingly, the de-attenuated correlations would be the most informative statistical data to record in the table since it shows the within-person variations. 9. In my view, the researchers have provided sufficient data to make the conclusion. The researchers made the conclusion after making similar observations in the 4 DGQs and BDHQs. 10. Two nutrients with very low correlation are the n-3 poly unsaturated fat and the alpha-carotene. Apart from the Pearson correlation coefficient, Rank correlation coefficient can be used to measure the degree to which any two nutrients vary during intake. Rank correlation coefficients, such as the Spearman’s and Kendall’s will show the relationship between the two nutrients. In addition, rank coefficients do not involve a lot of calculations. I would also recommend that the data be presented in graphs and charts in order to capture the differences in the dietary intake. Read More