Background The workplace is an important determinant of health that people are exposed to for the first-time during adolescence/early adulthood. This study investigates how diet, exercise and sleep changes as people aged 16-30 transition into work and whether this varies for different individuals and job characteristics.

Methods Multilevel linear regression models were used to assess changes in daily fruit and vegetable intake, usual hours of sleep, and exercise behaviours among 3,303 UK Household Longitudinal Study (UKHLS) participants aged 16-30 years who completed the adult surveys and started work for the first time between 2015 and 2023. Interrupted time series models assessed underlying behavioural trends in the period before and after starting work, an instantaneous effect of starting work, and an individual-level random intercept. Stratified analyses examined differences by selected demographic and employment characteristics. All analyses were done in R v.4.3.2.

Results Changes in fruit and vegetable were unrelated to starting work (fruit: β=-0.03 [95% CI: -0.11 to 0.05] portions/day and vegetable β= 0.02 [95% CI: -0.07 to 0.12] portions/day). No differences were found between men and women or between job characteristics. Sleep decreased immediately after starting work (β=-9.6 [95% CI: -16.8 to -2.4], hours/night), particularly among men (β=-13.8 [95% CI: -24.6 to -3.0] minutes/night) vs women (β=-6.0[95% CI: -15.0 to 3.0] minutes/night) but was otherwise stable over time. Exercise, measured in minutes of Metabolic Equivalent Tasks (MET) (a measure of exercise intensity) per day, increased immediately after starting work (β=113.30, [95% CI: 78.97 to 147.64], MET-min/day), but decreased over time afterwards (β=-26.70, [95% CI: -40.83 to -12.57] MET-min/day/year). The increase in exercise was greater in men (β=181.65, [95% CI: 123.67 to 239.64] MET-min/day) than in women (β= 62.68, [95% CI: 21.74 to 103.61] MET-min/day). Starting an office-based job and jobs away from home was associated with increased exercise (office: β=110.94 [95% CI: 71.82,150.06] MET-min/day and jobs away from home: β=241.85 [95% CI: 98.60 to 385.09] MET-min/day), whereas working from home was associated with an initial decrease in exercise after starting β=-126.32, [95% CI: -232.34 to -20.30] MET-min/day).

Conclusion To our knowledge, ours is the first study to examine how diet, exercise and sleep behaviours in young adults change as they enter the workforce. Starting work is associated with decreased sleep time and increased physical activity, and there are differences based on gender and job characteristics. Further research is needed to understand the mechanisms behind how these behaviours change during the transition to employment. Employers should consider how the workplace environment can be utilized to promote healthy behaviour interventions.