Abstract
Objectives There is increasing evidence to suggest an association between sleep and diet. The aim of the present study was to examine the association between sleep duration and fruit/vegetable (FV) intakes and their associated biomarkers in UK adults.
Design Cross-sectional.
Setting Data from The National Diet and Nutrition Survey.
Participants 1612 adults aged 19-65 years were included, pregnant/breast ceding women were excluded from the analyses.
Outcome measures Sleep duration was assessed by self-report, and diet was assessed by 4-day food diaries, disaggregation of foods containing FV into their components was conducted to determine total FV intakes. Sleep duration was divided into: short (<7 hours/day), reference (7-8 hours/day) and long (>8 hours/day) sleep periods. Multiple regression adjusting for confounders was used for analyses where sleep duration was the exposure and FV intakes and their associated biomarkers were the outcomes. Restricted cubic spline models were developed to explore potential non-linear associations.
Results In adjusted models, long sleepers (LS) consumed on average 28 (95% CI-50 to 6, p=0.01) g/day less of total FV compared to reference sleepers (RS), whereas short sleepers (SS) consumed 24 g/day less (95% CI-42 to 6, p=0.006) and had lower levels of FV biomarkers (total carotenoids, (3-carotene and lycopene) compared to RS. Restricted cubic spline models showed that the association between sleep duration and FV intakes was non-linear (p<0.001) with RS having the highest intakes compared to SS and LS. The associations between sleep duration and plasma total carotenoids (p=0.0035), plasma vitamin C (p=0.009) and lycopene (p<0.001) were non-linear with RS having the highest levels.
Conclusions These findings show a link between sleep duration and FV consumption. This may have important implications for lifestyle and behavioural change policy.