Background Rotational instability due to anterior cruciate ligament (ACL) deficiency or reconstruction remains an essential issue of this pathology. Despite knee rotation is verified in the transverse plane, few biomechanical studies have focused on this plane when assessing ACL function.

Objective The aims of this study were to analyse the transverse plane kinetics and kinematics of ACL deficient knees under torsional weight-bearing conditions in order to detect dynamic rotational instability, and secondly, to find a kinetic pattern of the ACL deficient knees under these conditions.

Design Laboratory in vivo study.

Methods Eight subjects with chronic ACL deficiency classified as non-copers (mean age 31 years; mean time since injury 2 years), and eight healthy recreational athletes as a control group (mean age 27 years) were recruited for this study. Each subject performed a crossover and a pivoting-vertical-jump. Kinematic and kinetic data were collected and analysed with Codamotion® and MATLAB® systems. Comparisons were made for peak rotation and torque in the transverse plane between ACL deficient subjects, their contralateral healthy knees and the control groups using ANOVA. Transverse plane torque curves analysis was made comparing ACLd and control groups using the T test at forty time intervals of the stance period. Half subjects in each group were tested twice for reliability purposes using the Intraclass Correlation Coefficient (ICC).

Results Both groups were comparable in demographic and anthropometric characteristics. Mean IKDC subjective score was 69 (range 56–81) in the ACLd group and 97 (range 92–100) in the control group (p<0.001).

Torque curves analysis showed a statistically significant dual behaviour in the crossover task, with initial lower values followed by greater values in the ACLd group compared to controls. In the pivoting-vertical-jump, values were lesser during the whole stance in the ACLd group with significant differences in the middle stance phase.

Peak tibial internal rotation in the crossover task was 19.8˚, 13.7˚ and 19.1˚ for the ACLd, contralateral and control groups respectively (p=0.176), while in the pivoting-vertical-jump values were 18.6˚, 13.8˚ and 18.1˚ respectively (p=0.297).

Peak internal rotation torque values in the crossover task were 257, 178.4 and 184.8 Nmm for the ACLd, contralateral and control groups respectively (p=0.5), while in the pivoting-vertical-jump values were 218.4, 152.2 and 238 Nmm respectively (p=0.288).

Reliability was good for all studied variables (ICC>0.75).

Conclusions Torque curve analysis identified a distinctive pattern comparing ACLd and control groups. The crossover showed a dual behaviour with an initial avoidance pattern followed by reinforcement in the last third of the turn. The pivoting-vertical-jump showed an avoidance pattern significant in a small portion of the middle stance. Despite the tendency towards an increase in the tibial internal rotation, a dynamic rotational instability was not detected.