Paul Cacolice

Defense Date


Graduation Date

Summer 8-8-2015


One-year Embargo

Submission Type


Degree Name



Rehabilitation Science


Rangos School of Health Sciences

Committee Chair

Christopher Carcia

Committee Member

Jason Scibek

Committee Member

Amy Phelps


Anterior Cruciate Ligament, Dorsiflexion, Fat Free Mass, H:Q Ratio, Margaria Kalmen, Triple Hop


Purpose: This study aimed to generate models predicting Ground Reaction Forces (GRFs), an established predictor of ACL injury incidence, from practical functional and clinical tests. Participants: Forty-two healthy, active college age individuals (21 females, age 20.667±1.461; 70.702±2.363cm; 82.202±7.606kg; 21 males, age 21.571±1.28; 65.524±1.874cm; 64.190±9.059kg) participated.

Methods and Materials: After assuring all participants met inclusion criteria and provided consent, lower extremity (LE) dominance was determined with drop landings. Individuals then had Fat Free Mass [FFM] determined from skinfolds, ankle joint dorsiflexion passive range of motion taken with a standard goniometer [DPROM], and performed the overhead deep squat test [ODS]. A warm-up on a bicycle ergometer then preceded determination of vertical [GRFz] and posterior ground reaction forces [GRFy] with five, signal-averaged LE drop landings from 35cm height onto a forceplate. Participants then performed the following tests in a counterbalanced order: Margaria-Kalamen [MK], Single Leg Triple Hop [SLTH], isometric peak force for lateral hip rotation [HipLR], knee flexion and knee extension. The knee flexion and extension peak force data was used to calculate a flexion:extension peak force ratio [H:Q] while GRFz and GRFy values were normalized to the participant’s FFM [nGRFz and nGRFy]. Stepwise linear regression models to predict the GRFs were calculated using FFM, DPROM, ODS, MK, SLTH, HipLR, H:Q and sex as the predictors. Alpha levels for all analyses were set a-priori at P≤ .05.

Results: Step-wise linear regression analysis indicated that a significant nGRFz model occurred utilizing all independent variables (Adjusted R2= .197, P= .048), but was most parsimonious with only SLTH and DPROM as predictor variables (Adjusted R2= .274; P=.001). Use of all eight-predictor variables for nGRFy also resulted in a statistically significant result (P= .001) but the most parsimonious model occurred with only H:Q, FFM and DPROM (Adjusted R2= .476; P< .001).

Conclusions: Two models significantly predicted GRFs from practical clinical measures and functional tests. One model predicted vertical ground reaction force from SLTH and DPROM, while one model predicted nGRFy from H:Q, FFM and DPROM. Clinical Relevance: If validated, a practical method of predicting nGRFy would be available to identify those at elevated ACL injury risk.