PMC

Individual and cumulative contribution of each factor to prostate cancer predictive accuracy. The area under the ROC curve (AUROC), as indicator of predictive accuracy for each variable in the risk model is shown. The left discontinuous (blue) line indicates the individual contribution of each variable, and the right continuous (red) line indicates the cumulative contribution, bottom to top. Environmental variables are sorted by increasing AUROC.

Distribution of prostate cancer cases and controls, and odds ratios for each decile of the genetic risk score. The left axis scale indicates the OR for prostate cancer according to deciles of points in the genetic score. The decile 1 (4.5–6.0 points) has been selected as reference category (OR = 1). The right axis scale indicates the proportion of cases and controls shown in bars for each decile. Parentheses include the points in the score.

Estimation of prostate cancer incidence in Spain by age (years) and risk score. Color lines indicate age-specific incidence rates of prostate cancer per 100 000 individuals in Spain according to risk score (RS), for a selection of values. The incidence curve for the average individual corresponds to RS = 1. The risk score can be calculated as RS = 2.47^(ERS-0.94) * 3.32^FH * 2.05^(GRS-6.98), where ERS is the points in the environmental score (average 0.94 in the population), FH is the presence of family history of prostate cancer (0 = no, 1 = yes) and GRS is the points in the genetic score (average 6.98 in the population).