Full text
| |
| Author(s): |
Lais Duarte Batista
Total Authors: 1
|
| Document type: | Master's Dissertation |
| Press: | São Paulo. |
| Institution: | Universidade de São Paulo (USP). Faculdade de Saúde Pública (FSP/CIR) |
| Defense date: | 2020-10-23 |
| Examining board members: |
Regina Mara Fisberg;
Karina Pfrimer;
Lilian Cuppari Valle
|
| Advisor: | Regina Mara Fisberg |
| Abstract | |
Introduction: Studies assessing the population\'s food intake using methods based on the individual\'s report are recurrent. However, the implausibility in the reported energy intake can compromise the validity of the associations between diet and health, impairing the interpretation of results. In this context, doubly labeled water (DLW) stands out as a biological recovery marker used to validate individuals\' energy intake estimates. Objectives: To evaluate the validity of predictive equations for estimating total energy expenditure (TEE) and the implausible report of energy intake in older adults. Methods: Data from 38 older people of both sexes, who had TEE measured by DLW were used. A database composed of 41 older adults with characteristics similar to the population of this study was used to model new predictive equations for determining TEE. To estimate energy intake, two 24-hour food records were applied on non-consecutive days. The implausibility in reporting energy intake was determined by comparing the ratio between mean reported energy intake and total energy expenditure, calculating the 95% confidence interval. The implausible report obtained by the reference method was compared with the predictive equations of Goldberg et al., Black, McCrory et al., Huang et al., and Rennie et al. Multiple linear regression was used to model the new predictive equations for TEE. Accuracy was assessed considering ± 10% of the ratio between TEE measured by DLW and estimated by the equations. Agreement between methods was assessed using intraclass correction coefficient and Bland-Altman plots. Sensitivity and specificity analyze were performed to test the performance of predictive equations in determining underreporting. Results: The average TEE by DLW was 2,656.7 (405.6) kcal/day for men and 2,168.9 (376.9) for women. Vinken et al. and both Dietary Reference Intake (DRIs) equations showed moderate to good agreement, while the developed models ranged from moderate to very good agreement compared to DLW. The accuracy was the same for the DRI equations and Vinken et al. (60.53%). The new equations developed in this study had accuracy ranging from 43.11% to 73.68% in the prediction of TEE in the elderly. The prevalence of underreporting and overreporting was 57.9% (n = 22) and 5.3% (n = 2), respectively. Black\'s equation showed the worst agreement while the method proposed by McCrory et al. the best agreement in classifying individuals in the three categories of implausible reporting. The McCrory et al. also had the best performance in the sensitivity and specificity analysis for detecting underreporting. Conclusion: The study showed a wide variation in the accuracy of predictive methods to determine implausible reporting, with none of the protocols showing excellent agreement compared to the classification obtained by the DLW. The results of this study highlight the importance of considering the implausible report of energy intake in studies that investigate the individuals\' dietary intake, to obtain more accurate and reliable data in studies that relate food and health. (AU) | |
| FAPESP's process: | 18/01991-4 - Implausibility in self-report of energy dietary intake in the population of São Paulo: comparison of consumption data and energy expenditure by means of a biological marker |
| Grantee: | Lais Duarte Batista |
| Support Opportunities: | Scholarships in Brazil - Master |