Association of medications with incident fractures in patients with rheumatoid arthritis
| Unadjusted HR (95% CI) | Adjusted HR* (95% CI) | |
| DMARD groups | ||
| Methotrexate monotherapy | Reference | Reference |
| TNFi | 0.79 (0.66 to 0.96) | 0.82 (0.67 to 1.00) |
| Non-TNFi bDMARDs | 1.03 (0.77 to 1.37) | 0.84 (0.61 to 1.16) |
| Other DMARDS | 0.88 (0.73 to 1.05) | 0.82 (0.69 to 0.99) |
| Glucocorticoid use | ||
| None | Reference | Reference |
| <7.5 mg/day for <3 months | 1.34 (0.72 to 2.51) | 1.23 (0.65 to 2.29) |
| <7.5 mg/day for ≥3 months | 1.59 (1.36 to 1.86) | 1.26 (1.07 to 1.48) |
| ≥7.5 mg/day for <3 months | 2.00 (1.20 to 3.35) | 1.66 (0.99 to 2.77) |
| ≥7.5 mg/day for ≥3 months | 2.46 (2.02 to 3.00) | 1.57 (1.27 to 1.94) |
| Statins | 1.01 (0.88 to 1.17) | 0.91 (0.78 to 1.06) |
| Antidepressants† | 2.36 (1.62 to 3.43) | 1.32 (1.13 to 1.56) |
| SSRIs | 1.76 (1.50 to 2.07) | 1.37 (1.15 to 1.63) |
| Others | 1.42 (1.13 to 1.77) | 1.03 (0.81 to 1.30) |
| PPIs | 1.36 (1.19 to 1.55) | 0.92 (0.80 to 1.06) |
| Opioids† | 3.78 (2.95 to 4.84) | 1.47 (1.27 to 1.71) |
| Weak opioids | 1.90 (1.65 to 2.18) | 1.37 (1.18 to 1.59) |
| Strong opioids | 2.66 (2.18 to 3.23) | 1.53 (1.24 to 1.88) |
| NSAIDs† | 1.07 (0.67 to 1.71) | 1.07 (0.94 to 1.23) |
| COX-2 inhibitors | 1.01 (0.83 to 1.23) | 0.99 (0.81 to 1.21) |
| Non-selective NSAIDs | 0.94 (0.82 to 1.08) | 1.13 (0.97 to 1.30) |
| Anticonvulsants | 1.47 (1.21 to 1.78) | 0.96 (0.78 to 1.17) |
| Antipsychotics | 1.55 (1.10 to 2.17) | 1.03 (0.73 to 1.47) |
| Benzodiazepines | 1.61 (1.27 to 2.05) | 0.94 (0.73 to 1.20) |
*Stratified by FRAX risk categories for MOF and adjusted for age, sex, ethnicity, RA duration, education level, insurance, rural residency, smoking, influenza vaccination, comorbidity index, BMI, HAQ, pain and patient global scores, prior osteoporosis diagnosis, use of osteoporosis specific medications, exercise, mental component score of SF-36, prior csDMARD and bDMARD counts, hospitalisation and calendar year.
†Opioid strength and antidepressant and NSAID types were assessed in a different model using the same covariables.
COX-2, cyclo-oxygenase-2; DMARD, disease-modifying antirheumatic drugs;NSAIDs, non-steroidal anti-inflammatory drugs;PPIs, proton pump inhibitors;RA, rheumatoid arthritis; SSRIs, selective serotonin reuptake inhibitors;TNFi, tumour necrosis factor-α inhibitor;bDMARDs, biological disease modifying antirheumatic drugs.