Patients (n=100, 38 male and 62 female, aged 41–63 years) were diagnosed with KOA in the First People’s Hospital of Yancheng City from January 2010 to December 2011. All patients provided their informed consent. The patients were randomly divided into the treatment and control groups according to the order of their diagnosis times. The patients in the treatment group received daily oral diclofenac sodium (Difene^®) and kinesitherapy. The patients in the control group received only daily oral diclofenac sodium. The treatment group contained 50 patients (20 males and 30 females, aged 44–62 years, mean 53.0±5.6). The numbers of patients having disease durations of <3 months, 3 months to 1 year, 1–2 years and 2–3 years were 9, 13, 18 and 10, respectively. The control group contained 50 patients (18 males and 32 females, aged 41–63 years, mean 52.3±7.0). The numbers of patients having disease durations of <3 months, 3 months to 1 year, 1–2 years and 2–3 years were 8, 15, 20 and 7, respectively. The patients in the two groups were comparable in gender distribution, age and disease duration without any statistically significant differences (P<0.05). The study was approved by the ethics committee of Yancheng Health Vocational and Technical College, Yancheng, China.

The knee osteoarthritis diagnostic criteria were as defined by the American Rheumatism Association (4): a) knee joint disease (pain duration ≥14 days within one month prior to treatment); b) X-ray showing osteophyte formation at the joint margins; c) clear, sticky synovial fluid with a leukocyte count <2,000 cells/ml; d) absence of synovial fluid index, age >40 years; e) knee morning stiffness <30 min; and f) crepitation during knee motion. Patients with a,b, or a,c,e,f, or a,d,e,f combinations of the above criteria were confirmed.

Inclusion criteria included: diagnostic criteria of KOA; not receiving other treatment methods or medication; age 40–70 years; and a disease classification of I-III by X-ray. Exclusion criteria included: significant stenosis of the joint space or bone bridge connection between the joints manifested as bony ankylosis; disease classification IV by X-ray; primary disease of the knee affecting joint structures; active gastrointestinal diseases and/or esophageal or peptic ulcer within 30 days of the trial; concurrent cardiovascular, cerebrovascular, liver, kidney or hematopoietic system diseases; mental illness; pregnancy and lactation; and known allergy to diclofenac.

The patients in the control group received oral diclofenac sodium (75 mg, twice daily). Patients in the treatment group received oral diclofenac sodium (75 mg, twice daily) and kinesitherapy. Kinesitherapy included knee joint flexion and extension (active exercise) followed by isometric contraction of the quadriceps femoris at angles of 0° and 90° from the knee joint. Contractions were held continuously for 10 sec followed by 10 sec of relaxation, repeated 10 times for one exercise circle. The patients also performed 5 circles at each angle, with 1 min between circles. Isometric exercises were performed in the sitting position to avoid the effects of weight-bearing on the knee joints. Patients carried out the exercises twice per day in the morning and afternoon, 4 days per week (5). The two groups received treatment for 4 consecutive weeks.

The joint function index refers to the sum of the evaluation scores for rest pain, activity pain, tenderness, swelling and joint function. Rest pain was scored as: painless; mild pain, no effect on life and work; moderate pain, effect on life and work; or severe pain, effect on sleep. The activity pain was scored as: painless; mild pain, pain during forced activity; moderate pain, pain during common activity; or severe pain, pain during light activity. Tenderness was scored as: no tenderness; mild tenderness, no effect on activity; moderate tenderness, frown; or severe tenderness, retreat. Swelling was scored as: no swelling; light swelling, limited to local sites of the joints; moderate swelling, in the joint range; or severe swelling, diffusing to the surrounding tissue. Joint function was scored as: normal; mild limitation, no effect on life and work; moderate limitation, effect on life and work; or severe limitation, sitting and/or lying position, unable to take care of themselves. For all parameters, the four possible scores were 0, 1, 2 and 3 (6).

Synovial fluid (1.5 ml) was extracted from the medial compartment of the knee prior to and following 4 weeks of treatment and was stored at −20°C until further analysis. An enzyme-linked immunosorbent assay (ELISA) was used to evaluate the TNF-α, hs-CRP and MMP-3 levels in the synovial fluid. The ELISA kits were acquired from Wuhan Boster Biological Engineering, Ltd. (Wuhan, China). The assays were performed according to the manufacturer’s instructions.

The therapeutic effect was determined using the Clinical Guideline of New Drugs for Traditional Chinese Medicine (7). The therapeutic effects were divided into four grades: clinically-controlled; significantly effective; effective; and ineffective.

SPSS 17.0 software was used for analysis. The independent Student’s t-test was used to compare the ages of the two groups and the χ² test was used to compare gender and disease duration difference. Efficacies were compared using the Mann-Whitney U test. The knee function index scores and cytokine levels prior to and following treatment were compared by repeat-measures analysis of variance. The above analyses were two-sided and P<0.05 was considered to indicate a statistically significant result.