PMC

Age distribution of starters from 1986 to 2007.

Finish rates from 1977 to 2007. Sex-specific finish rates were not available until 1986, so overall finish rates are displayed from 1977 through 1985.

Total number of starters from 1974 to 2007, and number of starters of each sex from 1986 to 2007. Sex-specific starting data are not available before 1986.

Frequency distribution of number of finishes for those individuals having finished the run from 1974 to 2007. Note that the vertical axis is on a log scale.

Mean ages of starters across the period of 1986 through 2007. Men are represented by open circles, and women are represented by closed circles. Brackets represent 1 SD and are displayed in only one direction for clarity.

Linear regressions defining the percentages of starters within different age groups across the period of 1986 through 2007. Solid lines represent conditions where the slopes of the linear regressions were significantly different (P < 0.05) from zero. Dashed lines are indicative of slopes that were not significantly different from zero.

Distribution of finish times for years 1986 through 2007. Men are represented by stippled bars; women are represented by solid bars. The small number of finishers between 30 and 32 h was because cutoff times were extended to 32 h in 1995 and 1998 because of heavy snow on the early portion of the course.

Mean finish times by year for top 5 finishers (upper panels) and all finishers (lower panels) in each age group for women (left side) and men (right side). Solid lines represent conditions where the slopes of the linear regressions were significantly different (P < 0.05) from zero. Dashed lines indicate the slopes of the linear regressions were not significantly different from zero. Top 5 data were considered when there were ≥10 finishers in the age group, and the all-finisher data were considered when there were ≥5 finishers in the age group. Lines traverse only the range of years for which data were considered.

Mean ages of top 5 overall men and women finishers from 1978 for men (open circles) and 1980 (first year with more than five women finishers) for women (closed circles). Brackets represent 1 SD and are displayed in only one direction for clarity. Solid line represents the linear trend for men (slope = 0.21, r² = 0.43, P < 0.001). The two broken lines represent linear trends for women across the time ranges of 1980 through 1989 (slope = −0.55, r² = 0.41, P = 0.045) and 1990 through 2007 (slope = 0.19, r² = 0.48, P = 0.0015).

Top 5 time spreads for men and women across time. Data are shown from 1978 for men and 1980 for women (first years with more than five finishers for either sex) with linear regressions across this time span shown by the solid lines. The slopes for both curves were not different from zero (P = 0.12 for men, P = 0.72 for women), but the top 5 time spread for women in 1980 was 7 percentage points higher (P < 0.05) than for men. The dashed line is the linear regression for women between 1990 and 2007, where time spread decreased by 0.75 percentage points per annum (P = 0.003).

Mean finish times as a function of age group (<30, 30–39, 40–49, 50–59, 60–69, and ≥70 yr) for men (open circles) and women (closed circles), considering the average for all finishers (solid lines) as well as the fastest finishes (broken lines) for each year. Data are for 1986 through 2007, plotted at the mean age of those considered within each age group. Brackets represent 1 SD and are displayed in only one direction for clarity. Data displayed for the ≥70 men represent a total of four finishes in 2 yr.

Mean finish times for the top 5 overall finishers across time (top panel) and differences in average finish times between the top 5 women and top 5 men as a percentage of the average time for the top 5 men (bottom panel). In the top panel, data are presented from 1978 for men and 1980 for women, which were the first years in which there were more than five finishers from either sex. The linear regression curve for the women was defined by the equation y = −0.062x + 144.9 (r² = 0.37, P = 0.0006); the curve for the men had a slope that was not significantly different from zero. In the bottom panel, the linear regression curve was defined by the equation y = −0.402x + 821.1 (r² = 0.43, P = 0.0002). Data before 1986 (when the course was less than a full 161 km) were included because other yearly variations in performance seemed to have greater effects on finish times than the modifications in the course.