Behaviours across a 24-hour day, including physical activity, sedentary time and sleep, are disrupted following cancer and contribute to cancer-related outcomes. This study describes the day-to-day 24-hour behaviour profiles of individuals with and without cancer, considering time since diagnosis and cancer types.
Seven days of accelerometer data from the UK Biobank (M±SDage=62.3±7.9 years; 56.4% female) were derived from machine learning models to assess the 24-hour behaviours in individuals with cancer (n=10 152; M±SDyears since diagnosis=7.4±6.1 years) compared with healthy (free of diseases) individuals (n=13 722). Diagnoses were identified using the International Classification of Disease codes within cancer registries. Bayesian compositional data analysis compared profiles between individuals with and without cancer, across time since diagnosis (<1 year, 1–5 years, >5 years) and 14 cancer types.
The least physically active profiles were observed for individuals within 1 year following cancer diagnosis and in cancers with poor prognoses. Compared with healthy individuals, those within 1 year following cancer diagnosis had 40 min/day less physical activity (light plus moderate-to-vigorous intensities), compensated by 40 min/day more inactive time (sedentary plus sleep periods). Differences also varied across cancer types, ranging from 22–75 min/day less physical activity and 22–75 min/day more inactive time, between individuals with cancers and healthy individuals. Cancers with poorer prognoses (eg, lung, gastrointestinal tract) had the least optimal profiles, whereas cancers with better prognoses (eg, prostate, skin) showed profiles closer to healthy individuals.
The 24-hour behaviour profiles differed by cancer history, prognosis and type. Supporting a healthy balance of behaviours, that can feasibly be achieved within a 24-hour day, should be considered for cancer survivors, particularly in the year after diagnosis and in poor prognosis cancers.