Figure 4

The relationship between hypertension, cholesterol, and diabetes (columns) and food-related metrics (rows). On the x-axis, we represent the relative food-consumption values compared to the average (set at 0), which are computed with Formula (9). On the y-axis, we represent the per-capita disease prevalence, which is computed with Formula (1): a value of 1 for a disease means that each resident takes, on average, 1 medication for that disease in the year. The dotted red line indicates the average disease prevalence across all areas. The shaded areas show 95% confidence intervals