Predicting Projected Frontal Area For Cycling In Women Competing At The 2004 Hawaiian Ironman Triathlon

Abstract

Purpose Previous studies have developed equations for predicting the projected frontal area (AP m2) of male cyclists (EJAP 85:358–366, 2001; EJAP 87:520–528, 2002), but similar analyses have never been performed for female cyclists. Such equations are needed for the creation of generalized models of cycling performance in women. Thus, the present study was designed to determine the best AP prediction equation in female triathletes from anthropometric, body position, and bicycle geometry variables. Methods 16 women (Mean±SD: 41.8±11.6 yrs, 59.7±4.4 kg body mass, 1.66±0.0.06 m body height) competing in the 2004 Hawaiian Ironman World Championship Triathlon volunteered to have projected frontal area determined from a single photograph. A digital photo was taken of the frontal plane of the cyclist (front view) while they sat on their bicycle mounted to a stationary trainer. The cyclists were instructed to place their feet parallel with the ground, use their aero handlebars while looking forward, and wear their typical race clothing and helmet for the digital photograph. Both total AP (body of cyclist + bicycle; 0.437±0.041 m2) and body AP (cyclist only; 0.315±0.029 m2) were predicted using body height (BH, m), body mass, shoulder width (0.37±0.02 m), trunk angle (TA, 19.5±4.2°), and seat tube angle (STA, 75.6±1.8°) as dependent variables. Both total and body AP were separately determined as the average of two separate measurements of the same digital photograph using digital photo analysis software (ImageJ 1.28u). Computed areas within each photograph were calibrated against a calibration frame (0.578 m × 0.578 m, or 0.334 m2) with known dimensions. The calibration frame was located in the field of view at the midline of the subject's torso. Standard step-forward multiple regression procedures were used to identify the single best prediction equation while a paired t-test was used to compare actual and predicted AP values (0.05 alpha level). Results The final prediction equation was as follows (n = 32 images): AP = 0.963 +0.122×C −0.0133×STA −0.00194×TA +0.238xBH (R2 = 0.926, SEE = ±0.021 m2, P <0.0001) Where C was binary coded ‘0’ and ‘1’ for predicting body and total AP respectively. Actual AP (Mean±SE: 0.376±0.013 m2) did not differ significantly (P = 0.687) from predicted AP (0.378±0.012 m2). Conclusions The accuracy of the AP prediction equation is similar to that reported previously for male cyclists. This equation may be useful when integrated into mathematical models for predicting time-trial bicycling performances of women triathletes and cyclists.