An examination of dietary variation across diverse population subgroups and development of frameworks and tools to assist future novel data collection and reuse in underdeveloped areas of nutritional science

Background: Population minority groups are underrepresented in health and nutrition research, limiting our understanding of how dietary habits influence health outcomes of such groups. There is a need to ensure that future dietary data collected from vulnerable groups are accurately obtained and existing data are reused to maximise impact. Objective: To understand dietary differences across population subgroups and develop services that address existing data gaps through reliable data collection and reuse. Methods: This thesis was conducted in three stages; review of current research into dietary intake and food choice of racial/ethnic groups globally; analysis of national survey data to understand meal patterns and diet quality across racial groups; development of services to facilitate further nutrition research through accurate novel data collection and data reuse. Results: Our review demonstrated the paucity of standardised food data across racial/ethnic groups. Existing data reported differences in food group and nutrient intake as well as in barriers to healthy food consumption across groups. Furthermore, using existing data from U.K. and U.S. national surveys, diet quality is not equal across racial groups. Asian groups reported highest diet quality in terms of nutritional adequacy (≥55/100), healthiness (≥47/100) and sustainability (≥19/42) compared to others. A shift to more plant-based foods would prove beneficial for all racial groups, but particularly White and Black groups who had lower such intake than Asian and Other groups; p<0.001. Meal patterns also varied across racial groups. Black groups consumed the lowest number of meals and snacks within a single day (2.55 eating occasions daily vs. 3.02-4.57 eating occasions by remaining groups) and did not report regular times of intake. Meal patterns were associated with diet quality (R2 adj=0.15, p<0.001) but further analysis is needed to determine the relationship between patterns of intake and weight status. Services supporting data reuse, data merging, and data collection were developed in this thesis to enhance knowledge of dietary intake of vulnerable populations in future work. A diet quality assessment tool was approved by experts in the dietary intake domain as a valuable method of supporting future data re-users to maintain rigor of original data and to only reuse data when appropriate to do so. A data mapping framework was developed to facilitate mapping and merging of existing datasets which provides researchers with larger sample sizes and more varied pool of participants, especially useful when examining diet and health across population groups. Applying this mapping framework to existing datasets was examined and demonstrated no significant impact on mean food group and nutrient intake and high correlations were reported across intakes (r=0.79-0.97; p<0.001) when original and mapped data were compared. To ensure a diverse range of ethnicities are adequately represented in nutritional analysis conducted in Ireland in the future, Foodbook24, a web-based 24-hour dietary recall tool, was expanded to account for ethnic food and language preferences and was assessed for use among diverse ethnicities in Ireland. When compared to traditional interviewer-led 24-hour dietary recalls, mean food group and nutrient intakes recorded via Foodbook24 were moderately correlated (r=0.52-0.96; p<0.001) and differences were mostly negligible when ethnic groups were analysed individually. Conclusion: This thesis has highlighted the need to consider future dietary intake data across diverse population subgroups as cohort-specific differences can inform future policy and effective intervention strategies aimed at mitigating inequalities in nutrition and health. The services developed as part of this thesis provide a robust platform to effectively collect and reuse dietary intake data of vulnerable groups who are currently underrepresented in nutrition research.