I see a Danish open-access study has just been published in the Canadian Medical Association Journal that may answer some of the questions. "Absolute 10-year risk of dementia by age, sex and APOE genotype: a population-based cohort study," http://www.cmaj.ca/content/190/35/E1033
Katrine L. Rasmussen, Anne Tybjærg-Hansen, Børge G. Nordestgaard and Ruth Frikke-Schmidt, CMAJ September 04, 2018 190 (35) E1033-E1041; DOI: https://doi.org/10.1503/cmaj.180066
The study aimed to determine the absolute 10-risk of dementia (including, separately and together, Alzheimer's, vascular, and unspecified), by age (60 to >80), sex, and APOE genotype. It included 104,537 participants (solely white European) from Copenhagen studies up to 2014. I won't copy the results here, as you can read the breakdowns in Figures 2 and 3 of the paper. There is no exact formula to calculate risk of factors other than genotype, as they present adjusted results. "Risk of dementia as a function of APOE genotype. Hazard ratios (HRs) were adjusted for age (time scale), sex, body mass index, hypertension, diabetes mellitus, smoking, alcohol consumption, physical inactivity, menopausal status and hormonal replacement therapy (only women), lipid-lowering therapy and education."
One rather surprising result is that e2/e4s had in this study a higher risk of vascular dementia than did e3/e4s (there is a brief explanation pertaining to lipid metabolism). The risk curves by age in Figure 2 are S-shaped rather than hyperbolic, as risk appears to taper off after a certain age (I was unable in my browser to read the legend in either the online HTML nor the downloaded PPT file, so I can't reproduce the numbers exactly).
I wonder if other readers could comment on whether this study is helpful for us to assess risk? I would like more information about the other risk factor adjustments such as physical inactivity, as the overall numbers seem a little lower than other published figures.