Values are number (%) unless otherwise stated.

Number of participants with missing data (the total number participants who have missing data, or who reported ‘do not know’ or ‘prefer not to answer’) for each characteristic is as follows: 0 for age, 263 for ethnicity, 150 for socioeconomic status, 756 for qualifications, 338 for BMI, 1704 for body fat %, 271 for height, 0 for excess MET-hours/wk, 64 for standing or walking job, 52 for manual job, 291 for smoking status, 44 for alcohol consumption, 676 for fruit and vegetable consumption, 655 for total red and processed meat consumption.

*Participants who reported their ethnicity as ‘White’, ‘British’, ‘Irish’ or ‘Any other white background’.

†We generated quintiles of socioeconmic status based on the Townsend deprivation index for the whole cohort (UK Biobank variable n_189_0_0).

‡Vocational qualifications defined as other professional qualification (eg. nursing or teaching)/ National Vocational Qualification or Higher National Diploma or Higher National Certificate) (touchscreen question number D12, UK Biobank variable n_6138_0_0).

§We preferentially used BMI derived from height and weight measured during the impedance measurement (UK Biobank variable n_23 104 _0_0_), but if missing used the body size measures (UK Biobank variable n_21 001 _0_0); both of these are direct measures of height and weight made on the same day at the assessment centre.

¶Body fat % (UK Biobank variable n_23 099 _0_0).

**Standing height (UK Biobank variable n_50_0_0).

††Excess MET-hours/wk estimated from the combination of reported walking, moderate and vigorous physical activity (for details see methods text).

‡‡Participants who reported their work ‘usually’ or ‘always’ involved walking or standing for most of the time (touchscreen question number D9B, UK Biobank variable n_806_0_0).

§§Participants who reported their work ‘usually’ or ‘always’ involved heavy manual or physical work for most of the time (touchscreen question number D9C, UK Biobank variable n_816_0_0).

¶¶Smoking status (UK Biobank variable n_20 116 _0_0).

***Participants who reported consuming alcohol three to four times per week or daily or almost daily (touchscreen question number A1, UK Biobank variable n_1558_0_0).

†††Total fruit and vegetable consumption is the sum of fresh fruit intake (touchscreen question number DT3, UK Biobank variable n_1309_0_0), cooked vegetable intake (touchscreen question numbers DT1, UK Biobank variable n_1289_0_0) and raw vegetable intake (touchscreen question number DT1, and UK Biobank variable n_1299_0_0). To sum the frequencies, ‘Less than one’ was coded as 0.5, and we coded 1 piece of fresh fruit as a serving and 2 tablespoons of vegetables as a serving.

‡‡‡Total red and processed meat consumption is the sum of processed meat (touchscreen question number DT8 and UK Biobank variable n_1349_0_0), beef (touchscreen question number DT7 and UK Biobank variable n_1369_0_0), lamb/mutton (touchscreen question number DT7A and UK Biobank variable n_1379_0_0), and pork (touchscreen question number DT7B and UK Biobank variable n_1389) intake. To sum the frequencies, we used the following coding: ‘Never’=0, ‘Less than once a week’=0.5, ‘Once a week’=1, ‘2–4 times a week’=3, ‘5–6 times a week’=5.5, ‘Once or more daily’=7.

BMI, body mass index; MET, metabolic equivalent.