Cardiometabolic health and risk of dementia and brain atrophy: a community-based prospective cohort study of 0.5 million adults in China

Figures

• 

  Fig. 1. Dementia and brain atrophy incidence rates, by age and sex. Sex-specific region-standardised incidence rates. Vertical lines represent 95% CIs.

• 

  Fig. 2. Associations of markers of prior cardiometabolic health with risk of dementia and brain atrophy. The hazard ratios (HRs) are shown as squares (not scaled proportional to the variance of the log risk) and 95% CIs are shown as horizontal lines. The chi-square and p-values are shown for heterogeneity between dementia subtypes. HRs were stratified by age-at-risk (5-year age groups), sex and region and adjusted for smoking, alcohol consumption, education, physical activity, healthy diet score, and BMI. Exposure is self-reported or screen-detected at recruitment or the first instance of the exposure identified by ICD-10 codes recorded at hospitalisation or in disease registries at least one year prior to incident dementia or brain atrophy. BMI, body mass index; IHD, ischaemic heart disease; TIA, transient ischaemic attack.

• 

  Fig. 3. Association of usual SBP with risk of all-cause dementia and brain atrophy. Each square represents the adjusted hazard ratio (HR) with the area inversely proportional to the variance of the category specific log HR. HRs are plotted on a floating absolute risk scale. Vertical lines indicate 95% CIs. The HRs are shown above each square and numbers of events below. Chi-squared tests were applied to the stratum-specific log HRs, using inverse-variance weights to assess evidence of a linear gradient across ordered strata. The regression dilution ratio (RDR = 0.50) for estimating usual SBP values was calculated using Rosner's regression method adjusting for region, sex, and 10-year age group. HRs are plotted against the mean usual SBP in each group, estimated by (B–A) x RDR + A, where A is the overall baseline mean SBP, and B is the baseline SBP group mean.