Login / Signup

Trends in food consumption according to the degree of food processing among the UK population over 11 years.

Mariana Ferreira MadrugaEurídice Martínez SteeleChristian J ReynoldsRenata Bertazzi LevyFernanda Rauber
Published in: The British journal of nutrition (2022)
Although ultra-processed foods represent more than half of the total energy consumed by the UK population, little is known about the trend in food consumption considering the degree of food processing. We evaluated the trends of the dietary share of foods categorised according to the NOVA classification in a historical series (2018-2019) among the UK population. Data were acquired from the NDNS, a survey that collects diet information through a 4-d food record. We used adjusted linear regression to estimate the dietary participation of NOVA groups and evaluated the linear trends over the years. From 2008 to 2019, we observed a significant increase in the energy share of culinary ingredients (from 3·7 to 4·9 % of the total energy consumed; P -trend = 0·001), especially for butter and oils; and reduction of processed foods (from 9·6 to 8·6 %; P -trend = 0·002), especially for beer and wine. Unprocessed or minimally processed foods (≅30 %, P -trend = 0·505) and ultra-processed foods (≅56 %, P -trend = 0·580) presented no significant change. However, changes in the consumption of some subgroups are noteworthy, such as the reduction in the energy share of red meat, sausages and other reconstituted meat products as well as the increase of fruits, ready meals, breakfast cereals, cookies, pastries, buns and cakes. Regarding the socio-demographic characteristics, no interaction was observed with the trend of the four NOVA groups. From 2008 to 2019 was observed a significant increase in culinary ingredients and a reduction in processed food. Furthermore, it sheds light on the high share of ultra-processed foods in the contemporary British diet.
Keyphrases
  • human health
  • physical activity
  • high resolution
  • machine learning
  • weight loss
  • cross sectional
  • risk assessment
  • healthcare
  • big data
  • deep learning
  • electronic health record