Table 4.
Adjusted Odds Ratios Relating Anticholinergic Use to Incident Recurrent Falls over the next 12 monthsa
| Model 1 | Model 2 | Model 3 | ||||
|---|---|---|---|---|---|---|
| Number With falls | OR (95 % CI) | Number With falls | OR (95 % CI) | Number With falls | OR (95 % CI) | |
| Any Anticholinergic use | ||||||
| No use | 7883 | 1.0 | 7883 | 1.0 | 7743 | 1.0 |
| Any use | 1938 | 1.86 (1.76, 1.97) | 1938 | 1.53 (1.45, 1.62) | 1903 | 1.51 (1.43, 1.60) |
| Number of Agents* | ||||||
| No use | 7883 | 1.0 | 7883 | 1.0 | 7743 | 1.0 |
| 1 | 1688 | 1.78 (1.67, 1.88) | 1688 | 1.48 (1.40, 1.57) | 1657 | 1.47 (1.38, 1.56) |
| 2+ | 250 | 2.84 (2.45, 3.28) | 250 | 2.04 (1.76, 2.36) | 246 | 2.00 (1.73, 2.32) |
| Duration* | ||||||
| No use | 7883 | 1.0 | 7883 | 1.0 | 7743 | 1.0 |
| <1 year | 495 | 1.82 (1.64, 2.01) | 495 | 1.52 (1.37, 1.68) | 485 | 1.50 (1.36, 1.66) |
| 1–3 years | 760 | 1.73 (1.59, 1.88) | 760 | 1.42 (1.31, 1.55) | 745 | 1.41 (1.29, 1.53) |
| >3 years | 683 | 2.10 (1.92, 2.29) | 683 | 1.70 (1.55, 1.86) | 673 | 1.68 (1.53, 1.84) |
| Specific Anticholinergic Classb | ||||||
| Antihistamine/Antiemetic/Antivertigo | 742 | 1.39 (1.28, 1.51) | 742 | 1.22 (1.12, 1.32) | 734 | 1.21 (1.11, 1.32) |
| Antidepressant/Antipsychotic | 811 | 2.26 (2.07, 2.46) | 811 | 1.83 (1.67, 1.99) | 796 | 1.81 (1.66, 1.97) |
| Gastrointestinal/Urinary Antimuscarinics/Miscellaneous | 607 | 1.81 (1.64, 1.99) | 607 | 1.49 (1.35, 1.64) | 591 | 1.47 (1.33, 1.62) |
Abbreviations: CI confidence intervals, OR odds ratio
*Trend test p-values all <0.0001
aOR and CI are estimated from GEE models using an unstructured correlation matrix, including baseline and year 3 medication exposure information, and recurrent falls in the following year. All models are adjusted by age using 5-year age intervals, and study component (clinical trial vs. observational study). Model 1 adjusted for linear age, diet modification and CaD trial randomization arms. Model 2 was further adjusted for race/ethnicity, body mass index, poor vision, arthritis, treated diabetes, low back pain, Alzheimer’s Disease diagnosis and/or use of medication, indications for use (Parkinson’s disease, urinary incontinence, depression, insomnia scale), self-reported health, alcohol intake, use of antiepileptics, and number of psychoactive medications. Model 3 was further adjusted for physical function, dizziness, current physical activity from walking. All adjustment variables are from baseline
bOdds ratios for type of anticholinergic medication are adjusted for simultaneous use of a different type