| dc.description.abstract | Lutein and zeaxanthin (L/Z) belong to carotenoids xanthophylls and are highly concentrated in the macula of the eye. The supplemental and dietary intake of L/Z have been associated with decreased risk and severity of age-related macular degeneration (AMD). Macular L/Z concentration, assessed as macular pigment optical density (MPOD), has been investigated as a proxy marker for AMD risk. In 2017 it was proposed that L/Z should have dietary intake targets considering their protective role. This proposal evaluated L/Z against the set of nine criteria developed by Lupton and colleagues to determine the strength of evidence to support intake targets. Criterion 3 refers to known food-concentration databases. The 2017 evaluation depended on the United States of America (US) food supply. Furthermore, the evidence to support dose-response relationships (criterion 6) largely relied on supplemental rather than dietary intake research. Therefore, the rationale for this thesis was to explore whether L/Z meet these criteria in countries other than the US.
Chapter 1 is a published narrative literature review appraising the evidence on MPOD response to dietary L/Z intake. There was minimal evidence of a dose-response relationship between dietary L/Z intake and MPOD. A large gap was that habitual dietary L/Z intake was not quantitatively monitored during intervention studies. Studies that did attempt measuring L/Z habitual intake used non-validated dietary intake tools.
Two additional gaps were identified related to determining the relationship between habitual dietary L/Z and MPOD. Firstly, the possible impact of blue light exposure from electronic device on MPOD status. Secondly, the paucity of data on food L/Z concentrations in food supplies (criterion 3), outside the USA (e.g. Australia).
These gaps are barriers to the valid measurement of habitual L/Z intake and relationships with MPOD. Therefore, the thesis aim was to determine how habitual dietary L/Z intake can best be validly and quantitatively measured.
Four studies were conducted to address this aim. Chapter 2 describes the development and validation process of two tools to quantitively monitor habitual dietary L/Z intake in healthy adults. Two screeners, with a recall timeframe of one and four weeks respectively were developed. L/Z intake reported from each screener was compared against multiple 24-hour diet recalls via Bland-Altman plot analysis to determine validity. Both screeners were significantly correlated (Spearman’s rank order, p<0.001) but returned poor validity compared with the 24-hour diet recalls (mean difference >0.3 mg/day). This indicated that participants were unable to report comparable L/Z intake between the tools; baby spinach contributed notably to discrepancies.
Chapter 3 describes the development and validation process of the Electronic Device Use Questionnaire (EDUQ). Healthy adults reported daily hours of device use using the EDUQ and multiple 24-hour diaries. EDUQ and diaries results were compared via Bland-Altman plot analysis; returning poor validity, indicating that participants were unable to report comparable device use.
Chapter 4 describes a cross-sectional study investigating whether MPOD was predicted by sex, age, estimation of electronic device use and dietary L/Z intake using the tools developed in this thesis. MPOD was not predicted by these variables in the 96 healthy Australian adults studied. Future research with more valid measurement tools should investigate this relationship further.
The food composition database in Australia reports only 26 food entries for L and none for Z. Analysis methods were not available for review. Chapter 5 describes the investigation of 12 extraction method variations on five Australian foods selected for known high L/Z concentration based on the US database. In this thesis, extraction refers to the isolation of L/Z from the food of interest for analysis of optimal recovery and measure of concentration per gram of fresh food. One variation was most optimal based on five foods for L, and four foods for Z. The L/Z concentration measured in these foods were notably higher or lower than that that reported in existing Australian and US composition databases.
Based on the work performed in this thesis, a dietary target for L/Z cannot yet be set with confidence, because the evidence available does not meet the nine criteria required in the framework proposed by Lupton and colleagues to determine dietary target values. The L/Z screener was unable to capture valid quantitative habitual dietary L/Z intake. A purposely developed questionnaire was not able to validly capture usual blue light exposure from electronic device to determine a relationship between electronic device use and MPOD. The purposely developed dietary L/Z screener, found to be invalid, indicated significant correlation between tools and simultaneous poor agreement on Bland-Altman analysis. This outcome suggests that results solely reliant on correlation statistics from prior research investigating the relationship between dietary L/Z and MPOD, or in the condition of AMD, should be interpreted with caution. Larger local L/Z food composition databases and valid tools for improved participant reporting of L/Z are needed to determine habitual L/Z intake and accurate relationship with MPOD. | en_GB |
