Table 5.
Multivariable logistic regression modeling: selected potential risk factors for osteonecrosis of the hip (subset N = 4134)
| Risk factor | Crude OR (95% CI) | Adjusteda OR (95% CI) |
|---|---|---|
| Drug exposures of interest (within the past 2 years) | ||
| Bisphosphonates | ||
| Intermittent | 6 (3.24, 11.15) | 1.7 (0.37, 7.72) |
| Exposed | 3.1 (1.17, 8.20) | 0.7 (0.11, 4.25) |
| Systemic corticosteroids | ||
| Intermittent | 4.2 (3.12, 5.58) | 3.1 (1.93, 4.95) |
| Exposed | 4.8 (2.84, 7.98) | 3 (1.37, 6.44) |
| Immunosuppressants | ||
| Intermittent | 22.4 (9.76, 51.54) | 6 (1.94, 18.38) |
| Exposed | 2.3 (0.45, 12.05) | 1.1 (0.07, 16.52) |
| Anti-infectives | ||
| Intermittent | 1.6 (1.26, 1.91) | 1.1 (0.79, 1.40) |
| Exposed | 1.7 (1.37, 2.22) | 1.2 (0.82, 1.65) |
| Statins | ||
| Intermittent | 0.7 (0.32, 1.36) | –b |
| Exposed | 0 (0) | –b |
| HRT (women only) | ||
| Intermittent | 1.1 (0.58, 2.27) | –c |
| Exposed | 1.7 (0.97, 3.15) | –c |
| Medical history in the 5 years prior | ||
| Hospitalization | 3.3 (2.61, 4.13) | 2 (1.43, 2.80) |
| Referral or specialist visit | 3.2 (2.53, 4.14) | 2.1 (1.50, 3.07) |
| Bone fracture | 6.5 (4.94, 8.47) | 5.8 (3.96, 8.56) |
| Any cancer, including hematological cancer | 3.2 (1.88, 5.55) | 2.8 (1.20, 6.31) |
| IBD | 10.5 (4.19, 26.50) | –b |
| Gout | 2.8 (1.47, 5.41) | 2.3 (0.85, 6.37) |
| Solid organ or bone transplantation | 24 (2.68, 214.68) | –b |
| Asthma | 1.8 (1.25, 2.57) | 1 (0.55, 1.73) |
| Renal failure or dialysis | 32.9 (7.31, 148.49) | –b |
| Congenital or acquired hip dislocation | 6 (0.85, 42.71) | –b |
| Diabetes mellitus | 0.8 (0.44, 1.36) | –b |
| Osteoporosis | 3.9 (2.23, 6.98) | 2.8 (0.93, 8.35) |
| Connective tissue disease | 5.6 (3.69, 8.64) | 2.5 (1.19, 5.39) |
| Osteoarthritis | 4.3 (3.35, 5.53) | 5 (3.51, 7.02) |
| Alcohol consumption | ||
| Missing | 0.9 (0.67, 1.33) | |
| Light drinker | 1.1 (0.78, 1.54) | |
| Moderate drinker | 1.4 (0.94, 2.22) | |
| Heavy/very heavy drinker | 2.7 (1.47, 5.03) | |
N = 601 cases and 3,533 controls
OR odds ratio; IBD inflammatory bowel disease; HRT hormone replacement therapy, Exposed 2+ prescriptions within 120 days in the past 2 years; Intermittent all other exposure scenarios
aThe final multivariable logistic regression model was adjusted for bisphosphonates, systemic corticosteroids, immunosuppressants, anti-infectives, hospitalization, referral or specialist visit, bone fracture, any cancer, gout, asthma, osteoporosis, connective tissue disease, and osteoarthritis
bVariables excluded from the final regression model based on either not reaching 1% overall prevalence or crude OR was not statistically significant
cHRT was excluded from the final regression model in order to retain the full sample (men and women)