Skip to main content
. 2020 Feb 12;368:m131. doi: 10.1136/bmj.m131

Fig 2.

Fig 2

Association of baseline biomedical characteristics with cardiovascular disease (CVD) mortality in UK Biobank and Health Survey for England/Scottish Health Surveys (HSE-SHS) cohort studies. Hazard ratios are adjusted for age and sex. Squares indicate hazard ratios and error bars denote 95% CI for relation of each characteristic with risk of death outcome. Ratio of hazard ratios (RHR) summarises between study differences (HSE-SHS is reference group) for effect estimates for each outcome. Distributions of glycated haemoglobin (HbA1c), C reactive protein, and high density lipoprotein (HDL) cholesterol were skewed, so they were log2 transformed and effect estimates reflect doubling for each biomarker. Elevated waist:hip ratio was denoted by ≥0.90 for men and ≥0.85 for women; obesity was indicated by body mass index ≥30. FEV1=forced expiratory volume in one second