Girls in the boat: Sex differences in rowing performance and participation

Abstract

Men outperform women in many athletic endeavors due to physiological and anatomical differences (e.g. larger and faster muscle); however, the observed sex differences in elite athletic performance are typically larger than expected, and may reflect sex-related differences in opportunity or incentives. As collegiate rowing in the United States has been largely incentivized for women over the last 20 years, but not men, the purpose of this study was to examine sex differences in elite rowing performance over that timeframe. Finishing times from grand finale races for collegiate championship on-water performances (n = 480) and junior indoor performances (n = 1,280) were compared between men and women across 20 years (1997-2016), weight classes (heavy vs. lightweight) and finishing place. Participation of the numbers of men and women rowers were also quantified across years. Men were faster than women across all finishing places, weight classes and years of competition and performance declined across finishing place for both men and women (P<0.001). Interestingly, the reduction in performance time across finishing place was greater (P<0.001) for collegiate men compared to women in the heavyweight division. This result is opposite to other sports (e.g. running and swimming), and to lightweight rowing in this study, which provides women fewer incentives than in heavyweight rowing. Correspondingly, participation in collegiate rowing has increased by ~113 women per year (P<0.001), with no change (P = 0.899) for collegiate men. These results indicate that increased participation and incentives within collegiate rowing for women vs. men contribute to sex differences in athletic performance.

Fig 1

NCAA and IRA performance times of men and women improved across competition year [* denotes a significant effect of time, p < 0.001]; however, women improved more than men [† denotes a significant time × sex interaction, P = 0.001] for the heavyweight (A) and lightweight class (B). Note that heavyweight women compete in the NCAA and lightweight men and women and heavyweight men in the IRA.

Fig 2

NCAA and IRA relative performance (% 1^st place) across finishing place for heavyweight (A) and lightweight class (B). The reduction in relative performance was greater for men than women [# denotes greater reduction in men, place × sex, P < 0.001] for the heavyweight class (A). Conversely, the reduction in relative performance was greater for women compared with men [† denotes greater reduction in women, place × sex, P < 0.001] for the lightweight class (B). Note that heavyweight women compete in the NCAA and lightweight men and women and heavyweight men in the IRA.

Fig 3

Ergometer performance times (“erg times”) of men and women in the Junior World Indoor Rowing Championships improved across competition year [* denotes significant time effect, P < 0.001]; although, women improved more than men [† denotes a significant time × sex interaction, P = 0.001] for the heavyweight (A) and lightweight class (B) during the World Indoor Championships.

Fig 4

In the Junior World Indoor Rowing Championships, relative ergometer performance (% 1^st place) reduced across finishing place (place effect, P < 0.001); further, women had greater reductions compared to men [† denotes a significant place × sex interaction, P = 0.001] for heavyweight (A) and lightweight class (B).

Fig 5. Sex difference in performance increased across finishing place for Junior World Indoor Championships [* denotes significant place effect, P < 0.001]; although, the sex difference was less for the heavyweight (open triangle) compared with the lightweight rowers (filled triangle) [† denotes a significant place × weight class interaction, P = 0.001] for top 16 place finishers.

Fig 6

Participation of men (filled circles) and women (open circles) competing in the NCAA regardless of weight class (A), in the heavyweight division of Junior Championships (B), and in the lightweight division of the Junior Championships (C). A: Participation among the heavyweight women in the NCAA increased linearly (y = 113.1x + b, r² = 0.847, P < 0.001); however, there was no significant regression among men in the NCAA. B: Participation increased among heavyweight men and women competing in the Junior Championships (y = 10.3x + b, r² = 0.908, P < 0.001; y = 11.7x + b, r² = 0.889, P < 0.001, respectively). C: Participation increased among lightweight men and women competing in the Junior Championships (y = 9.9x + b, r² = 0.861, P < 0.001; y = 7.4x + b, r² = 0.714, P < 0.001, respectively). (y = total number of participants, x = year of competition, b = constant).