Can internal load associate with adaptation? A comparison of 1v1 and 3v3 basketball training in physically inactive female college students

This study compared internal-load profiles between 1v1 and 3v3 basketball training and examined whether accumulated internal load was associated with changes in estimated maximal oxygen uptake (VO₂max) in physically inactive female college students. The longitudinal estimated VO₂max analysis included 63 randomized participants allocated to 1v1 basketball training, 3v3 basketball training, or a non-exercising control condition. Internal-load and dose-response analyses were restricted to the 42 exercisers because the control group had no training-load exposure. Participants in the exercise groups completed three supervised 60-min sessions per week for 8 weeks. Internal load was quantified using accumulated session-RPE (ΣsRPE) and mean peak heart rate (HRpeak), with CR10 expressed as the algebraic equivalent of ΣsRPE divided by the fixed total training duration. Compared with 3v3, 1v1 produced higher ΣsRPE (12,174 ± 203 vs. 11,351 ± 229 AU; mean difference, 823 AU; 95% CI, 688 to 958; p < 0.001; Cohen’s d = 3.80). Mean HRpeak was also higher in 1v1 than in 3v3 (178.33 ± 0.83 vs. 177.49 ± 0.50 bpm; mean difference, 0.83 bpm; 95% CI, 0.40 to 1.26; p < 0.001; Cohen’s d = 1.22), although the absolute difference was physiologically small. Estimated VO₂max increased in 1v1 by 1.97 mL kg −1 min −1 (95% CI, 1.40 to 2.55; p < 0.001) and in 3v3 by 2.45 mL kg −1 min −1 (95% CI, 1.60 to 3.31; p < 0.001), whereas the control group showed no clear change (0.07 mL kg −1 min −1 ; 95% CI, -0.09 to 0.24; p = 0.348). The group difference in pre-to-post change was statistically clear (F(2,60) = 18.93, p < 0.001, η² = 0.387), with greater improvements in both exercise groups than in control and no clear difference between 1v1 and 3v3. Among exercisers, ΔVO₂max was not clearly associated with ΣsRPE ( r = -0.050, 95% CI, -0.348 to 0.258; p = 0.755) or mean HRpeak ( r = -0.130, 95% CI, -0.418 to 0.181; p = 0.411). Neither association met the prespecified two-sided α = 0.05 criterion, and both 95% confidence intervals included the null value ( r = 0). These findings indicate that both basketball formats improved estimated aerobic fitness compared with control, while 1v1 imposed greater global internal load than 3v3. In this sample, and using the available global load indices, no statistically clear dose–response association was detected. Registration: The study protocol was registered in the ISRCTN (24697262).