Open Access Study protocol

Strength Training for Arthritis Trial (START): design and rationale

Stephen P Messier1103*, Shannon L Mihalko1, Daniel P Beavers2, Barbara J Nicklas13, Paul DeVita4, J Jeffery Carr5, David J Hunter6, Jeff D Williamson3, Kim L Bennell7, Ali Guermazi8, Mary Lyles3 and Richard F Loeser19

Author Affiliations

1 Department of Health and Exercise Science, Wake Forest University, Winston-Salem, NC 27109, USA

2 Department of Biostatistical Sciences, Wake Forest School of Medicine, Winston-Salem, NC, USA

3 Section on Gerontology and Geriatric Medicine, Wake Forest School of Medicine, Winston-Salem, NC, USA

4 Department of Exercise and Sport Science, East Carolina University, Greenville, NC, USA

5 Department of Radiology, Vanderbilt University Medical Center, Nashville, TN, USA

6 Rheumatology Department, Kolling Institute, University of Sydney, Sydney, Australia

7 Department of Physiotherapy, The University of Melbourne, Parkville, VIC, Australia

8 Department of Radiology, Boston University School of Medicine, Boston, USA

9 Section on Molecular Medicine, Wake Forest School of Medicine, Winston-Salem, NC, USA

10 Department of Rheumatology and Immunology, Wake Forest School of Medicine, Winston-Salem, NC, USA

For all author emails, please log on.

BMC Musculoskeletal Disorders 2013, 14:208  doi:10.1186/1471-2474-14-208

Published: 15 July 2013



Muscle loss and fat gain contribute to the disability, pain, and morbidity associated with knee osteoarthritis (OA), and thigh muscle weakness is an independent and modifiable risk factor for it. However, while all published treatment guidelines recommend muscle strengthening exercise to combat loss of muscle mass and strength in knee OA patients, previous strength training studies either used intensities or loads below recommended levels for healthy adults or were generally short, lasting only 6 to 24 weeks. The efficacy of high-intensity strength training in improving OA symptoms, slowing progression, and affecting the underlying mechanisms has not been examined due to the unsubstantiated belief that it might exacerbate symptoms. We hypothesize that in addition to short-term clinical benefits, combining greater duration with high-intensity strength training will alter thigh composition sufficiently to attain long-term reductions in knee-joint forces, lower pain levels, decrease inflammatory cytokines, and slow OA progression.


This is an assessor-blind, randomized controlled trial. The study population consists of 372 older (age ≥ 55 yrs) ambulatory, community-dwelling persons with: (1) mild-to-moderate medial tibiofemoral OA (Kellgren-Lawrence (KL) = 2 or 3); (2) knee neutral or varus aligned knee ( -2° valgus ≤ angle ≤ 10° varus); (3) 20 kg.m-2 ≥ BMI ≤ 45 kg.m-2; and (3) no participation in a formal strength-training program for more than 30 minutes per week within the past 6 months. Participants are randomized to one of 3 groups: high-intensity strength training (75-90% 1Repetition Maximum (1RM)); low-intensity strength training (30-40%1RM); or healthy living education. The primary clinical aim is to compare the interventions’ effects on knee pain, and the primary mechanistic aim is to compare their effects on knee-joint compressive forces during walking, a mechanism that affects the OA disease pathway. Secondary aims will compare the interventions’ effects on additional clinical measures of disease severity (e.g., function, mobility); disease progression measured by x-ray; thigh muscle and fat volume, measured by computed tomography (CT); components of thigh muscle function, including hip abductor strength and quadriceps strength, and power; additional measures of knee-joint loading; inflammatory and OA biomarkers; and health-related quality of life.


Test-retest reliability for the thigh CT scan was: total thigh volume, intra-class correlation coefficients (ICC) = 0.99; total fat volume, ICC = 0.99, and total muscle volume, ICC = 0.99. ICC for both isokinetic concentric knee flexion and extension strength was 0.93, and for hip-abductor concentric strength was 0.99. The reliability of our 1RM testing was: leg press, ICC = 0.95; leg curl, ICC = 0.99; and leg extension, ICC = 0.98. Results of this trial will provide critically needed guidance for clinicians in a variety of health professions who prescribe and oversee treatment and prevention of OA-related complications. Given the prevalence and impact of OA and the widespread availability of this intervention, assessing the efficacy of optimal strength training has the potential for immediate and vital clinical impact.

Trial registration, NCT01489462