Effect of bisphosphonate use on radiographic and clinical measures of knee osteoarthritis over 4 years of observation
| Non-users Mean change (SD)* n=258 | Bisphosphonate users Mean change (SD)* n=55 | Effect size (95% CI) (adjusted for age and BMI)† | p Value | |
|---|---|---|---|---|
| Pain score (numerical rating score) | ||||
| Year 1 | −0.9 (2.8) | −1.1 (2.9) | −0.23 (−0.87 to 0.42) | 0.491 |
| Year 2 | −1.0 (2.9) | −2.3 (2.5) | −0.97 (−1.55 to −0.38) | 0.001 |
| Year 3 | −0.9 (3.2) | −2.2 (2.7) | −1.15 (−1.94 to −0.36) | 0.004 |
| Year 4 | −1.1 (3.2) | −1.3 (2.8) | 0.30 (−0.58 to 1.19) | 0.503 |
| WOMAC pain score | ||||
| Year 1 | −0.7 (4.3) | −1.5 (3.8) | −0.44 (−1.30 to 0.42) | 0.316 |
| Year 2 | −0.9 (4.2) | −2.2 (3.8) | −0.72 (−1.67 to 0.23) | 0.137 |
| Year 3 | −0.8 (4.9) | −2.0 (3.8) | −0.34 (−1.33 to 0.64) | 0.492 |
| Year 4 | −1.3 (4.5) | −1.2 (4.0) | 0.69 (−0.54 to 1.92) | 0.271 |
| WOMAC disability score | ||||
| Year 1 | −2.3 (12.1) | −3.7 (11.4) | −1.62 (−5.08 to 1.85) | 0.361 |
| Year 2 | −3.1 (12.3) | −7.6 (11.8) | −2.85 (−6.30 to 0.61) | 0.107 |
| Year 3 | −3.2 (13.9) | −5.8 (14.0) | −2.29 (−5.70 to 1.12) | 0.188 |
| Year 4 | −4.5 (13.9) | −6.7 (10.8) | 0.05 (−3.85 to 3.95) | 0.982 |
| WOMAC stiffness score | ||||
| Year 1 | −0.3 (2.0) | −0.4 (1.7) | −0.06 (−0.46 to 0.34) | 0.773 |
| Year 2 | −0.4 (2.1) | −1.2 (1.7) | −0.12 (−0.51 to 0.27) | 0.552 |
| Year 3 | −0.5 (2.1) | −1.1 (1.8) | −0.18 (−0.61 to 0.25) | 0.417 |
| Year 4 | −0.6 (2.2) | −1.0 (1.5) | −0.24 (−0.75 to 0.27) | 0.358 |
| Change in JSW | ||||
| Year 1 | −0.15 (0.63) | −0.09 (0.4) | 0.13 (−0.11 to 0.38) | 0.292 |
| Year 2 | −0.35 (0.73) | −0.12 (0.52) | 0.20 (−0.07 to 0.48) | 0.144 |
| Year 3 | −0.50 (0.78) | −0.24 (0.61) | 0.28 (−0.04 to 0.59) | 0.083 |
| Year 4 | −0.51 (0.83) | −0.29 (0.75) | 0.35 (−0.01 to 0.70) | 0.057 |
| Radiological progression (n (%)) | ||||
| Year 1 | 45 (18.7) | 6 (12.0) | 0.59 (0.30 to 1.13) | 0.111 |
| Year 2 | 70 (31.8) | 11 (21.6) | 0.61 (0.36 to 1.03) | 0.063 |
| Year 3 | 82 (39.2) | 11 (22.9) | 0.63 (0.41 to 0.98) | 0.041 |
| Year 4 | 82 (40.6) | 13 (28.3) | 0.66 (0.43 to 1.01) | 0.057 |
| Clinical progression (%) | ||||
| Year 1 | 17 (25.0) | 3 (6.0) | 0.65 (0.25 to 1.72) | 0.385 |
| Year 2 | 25 (11.6) | 5 (9.8) | 0.70 (0.34 to 1.44) | 0.328 |
| Year 3 | 39 (19.4) | 6 (12.5) | 0.75 (0.42 to 1.33) | 0.323 |
| Year 4 | 41 (20.9) | 9 (19.6) | 0.80 (0.43 to 1.48) | 0.474 |
| Joint replacement (n (%)) | ||||
| Years 1–4 | 16 (6) | 2 (3.6) | 0.73 (0.15 to 3.48) | 0.688 |
The effect size is equivalent to differences in the unit on the y-axis for continuous outcomes, using the β coefficient (eg, mm JSW for x-ray), and equivalent to ratios of the differences for binary outcomes, using incident rate ratios.
Statistically significant p-values are indicated by bold font.
Numeric rating scale change and WOMAC pain change further adjusted for analgesic use and baseline pain score.
*Mean change is from baseline unless otherwise specified.
†Effect sizes are β coefficients, except when outcomes are binary (radiological and clinical progression, joint replacement) where effect sizes are incidence rate ratio.
BMI, body mass index; JSW, joint space width; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index.