Lettuce be happy: A longitudinal UK study on the relationship between fruit and vegetable consumption and well-being

Neel Ocean*, Peter Howley, Jonathan Ensor

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review


RATIONALE: While the role of diet in influencing physical health is now well-established, some recent research suggests that increased consumption of fruits and vegetables could play a role in enhancing mental well-being. A limitation with much of this existing research is its reliance on cross-sectional correlations, convenience samples, and/or lack of adequate controls.

OBJECTIVE: We aim to add to the emerging literature on the relationship between fruit and vegetable consumption and well-being by using longitudinal data from a study in the United Kingdom (UK).

METHOD: We employ panel data analytical techniques on three waves collected between 2010 and 2017 (i.e., following the same individuals over time) in the UK Household Longitudinal Survey. We also control for time-variant confounders such as diet, health, and lifestyle behaviours.

RESULTS: Fixed effects regressions show that mental well-being (GHQ-12) responds in a dose-response fashion to increases in both the quantity and the frequency of fruit and vegetables consumed. This relationship is robust to the use of subjective well-being (life satisfaction) instead of mental well-being. We also document a hump-shaped relationship between fruit and vegetable consumption and age.

CONCLUSION: Our findings provide further evidence that persuading people to consume more fruits and vegetables may not only benefit their physical health in the long-run, but also their mental well-being in the short-run.

Original languageEnglish
Pages (from-to)335-345
Number of pages11
JournalSocial Science & Medicine
Early online date7 Jan 2019
Publication statusPublished - 1 Feb 2019

Bibliographical note

© 2019 The Authors.


  • Diet
  • Fixed-effects
  • Fruit and vegetables
  • GHQ-12
  • Panel data
  • UK
  • Well-being

Cite this