Chapter 14 457 of the mid-point age of each age group versus % any gray and of % total or complete gray. As before, these plots are clearly not linear, but display a distinct curvature. Data from all three hair color types of the Keogh and Walsh study were combined and the data of men and women in the Copenhagen study were combined. The Keogh and Walsh study went down to age 25, while the Copenhagen study stopped at age 35 (30 to 39). On the other hand, the Keogh and Walsh study used few subjects at age 55 and above, whereas the Copenhagen study used many more subjects in the higher age groups. So, where possible, the Keogh and Walsh data were used for ages 20–30 and the Copenhagen study for ages 60–70. Means of both studies were averaged for ages 35–55. Table 2 summarizes the analysis. Table 2. Percent little gray, % moderate gray, and % total gray at different ages for female Caucasians, calculated from data of Keogh & Walsh6 and Schnohr et al.9 Age % Little Gray % Moderate Gray % Total Gray(Eqn) % Any Gray(Eqn) 20 0 0 0 0 25 6.3 0* 0 6.3 30 22.5 0.6* 0 23.1 35 43.5 3.7 0.5 42.8 40 52.3 9.3 2.7 61.4 45 54.7 16.4 6.0 76.2 50 52.2 22.7 10.3 86.0 55 46.2 30.1 15.7 90.9 60 38.0 33.7 22.2 92.4 65 29.1 36.8 29.8 93.3 70 20.8 37.5 38.5 98.0 % Little Gray is from a cubic equation for the Schnohr data of ages 35–70 and for ages 20–30 by subtraction. For the cubic model R2 = 0.99, p = 0.0015 and the root mean square error = 2.157 % Moderate Gray was obtained by subtraction of % Total Gray + % Little Gray from % Any Gray *this data point was from linear regression analysis and extrapolation from ages 35–70 % Total Gray is from a quadratic model of data combined from the Keogh and Walsh and Copenhagen studies where R2 = 0.994, p 0.0001 and root mean square error is 1.167 % Any Gray is from a quadratic model from data of Ke- ogh and Walsh and the Copenhagen studies combined where R2 = 0.996 and p 0.0001 and the root mean square error is 2.8799.