The most economically deprived people tend to have the poorest quality diets^{(Reference Darmon and Drewnowski1)}. However, healthier diets can be achieved at low cost or within low-income households^{(Reference Marty, Dubois and Gaubard2)}. This research sought to identify exemplars to facilitate improving the diet quality of people living in low-income households.

The National Diet and Nutrition Survey (NDNS)⁽³⁾ dataset collected April 2012 to March 2016 (survey years 5–8)⁽³⁾ was searched and individuals (n = 203) over 16 years with the highest Diet Quality Index (DQI) scores^{(Reference Craig, Barton and Horgan4)} and in the lowest two income quintiles identified as potential exemplars. Attempts were made to contact them; in-depth, semi-structured qualitative interviews were conducted with 17 participants (14 female, ≥25 years). Interview topics included: attitudes to food and eating; food purchasing, preparation and eating; food knowledge; food consumption environments, and interviewees’ socio-cultural context. Interviews were transcribed fully and coded axially according to the interview topics. Coding was done manually in NVivo.

Interviewees, who were mainly responsible for their households’ food shopping and preparation, reported preparing meals that were “pretty simple” (male, >65, retired, living with partner) consisting of grilled meat or fish with several portions of vegetables, pasta and rice dishes, stews, soups and salads. “I just love fresh vegetables and we've usually got plates heaving in vegetables with a bit of dead animal on the side” (female, >65, retired, living with partner). Interviewees’ food consumption behaviours were habitual: using the same stores, even when there were multiple food shops nearby; buying and cooking the same foods. Choice of food shops was influenced by convenience, though price and range were also important. Interviewees did not report seeking out nutritional information from food labels, nor educational materials, and indicated that food preparation knowledge was acquired from family, friends and school lessons.

Interviewees’ healthy diets (high DQI scores) appear to be the result of them eating simple meals prepared at home and making minimal use of restaurants and takeaways. However, although all interviewees were in the lowest two income quintiles (in the NDNS data), 12 owned their home, either outright or with a mortgage. Thus, they are unlikely to be economically deprived and, for social and cultural reasons, are unlikely to be dietary exemplars for people who are.

This study provides further evidence of the need for a more nuanced and holistic approach to healthy eating policy. The food consumption habits of one socio-cultural group cannot be generalised easily to others, and income alone is unlikely to be able to differentiate them.

As food consumption patterns developed during childhood were reported to be important for the healthy eating behaviours of our interviewees, further work to embed food preparation skills into the curriculum from early years should be explored rather than seeking additional data.