Figure 1 from Adiposity Status Close to Diagnosis and Its Association with Prostate Cancer Survival in the UK Biobank

Data collection time frames for adiposity data – study design. The baseline assessment visit, in which adiposity measurements were taken, occurred between 2006 and 2010 (instance 0). Adiposity measurements including body fat percentage by bioelectrical impedance were also taken from some individuals in the first repeat assessment visit that occurred between 2012 and 2013 (instance 1) and in the first imaging visit after 2014 (instance 2). The pre-diagnosis analyses (blue boxes) included men with incident prostate cancer who (1) either had baseline or (2) follow-up adiposity data (from any repeat assessment) collected up to 2 years before diagnosis. The post-diagnosis analyses (gray boxes) include (1) men with prevalent prostate cancer at study entry who had baseline data collected up to 5 years after diagnosis or (2) men with incident prostate cancer who had follow-up data collected up to 5 years after diagnosis (from any repeat assessment). Green boxes indicate the number of men with prostate cancer who participated in each assessment visit. Deaths: all-cause (prostate cancer specific). Each timeframe/period (a–f) excludes individuals from previous time frames/periods (a–f) as appropriate (no overlap of individuals in the final analytic sample for each adiposity index). The final analytic sample of the present study is those who had adiposity data on BMI, waist and hip circumference, and waist-to-hip ratio and complete data on the covariates of the main model [age of diagnosis, year of diagnosis, smoking status, physical activity, sedentary activities (sum of time spent watching TV, using a computer screen, or driving in hours/day), Townsend deprivation index (quintiles), and alcohol intake frequency. The person years of observation in the pre- or post-diagnosis analysis were 38,551; in the post-diagnosis analysis, they were 23,937, and in the pre-diagnosis analysis, they were 14,614 (rounded to the nearest whole number)].