| Baseline ET | Baseline control | RR (95% CI), p value | |
| Incident parkinsonism | 12 (5.8%) of 207* | 56 (1.6%) of 3606† | Unadjusted 3.68 (1.97–6.87), <0.001 |
| Adjusted‡ 3.47 (1.82–6.59), <0.001 | |||
| Incident PD§ | 6 (3.0%) of 201 | 24 (0.7%) of 3574 | Unadjusted 4.34 (1.77–10.62), 0.001 |
| Adjusted¶ 4.27 (1.72–10.61), 0.002 | |||
| Incident drug induced parkinsonism** | 4 (2.0%) of 199 | 18 (0.5%) of 3568 | Unadjusted 3.88 (1.31–11.48), 0.014 |
| Adjusted†† 3.06 (1.03–9.09), 0.04 | |||
| Incident non-PD parkinsonism‡‡ | 2 (1.0%) of 197§§ | 14 (0.4%) of 3564 | Unadjusted 2.54 (0.58–11.17), 0.22 |
| Adjusted¶¶ 1.90 (0.43–8.45), 0.40 |
*10 cases were diagnosed with parkinsonism based on a direct examination and two by review of medical information (medical records from general practitioners, inpatient hospitalisations and neurological specialists as well as death certificate diagnoses).
†52 cases were diagnosed with parkinsonism based on a direct examination and four by review of medical information (medical records from general practitioners, inpatient hospitalisations and neurological specialists as well as death certificate diagnoses).
‡Analysis adjusting for age, educational level and current smoker. These covariates were chosen because in univariate analyses, age and educational level were associated with baseline ET, and age and current smoker were associated with incident parkinsonism.
§We excluded participants who developed drug induced parkinsonism (n = 22) or non-PD parkinsonism (n = 16).
¶Analysis adjusting for age, gender and educational level. These covariates were chosen because in univariate analyses, age and educational level were associated with baseline ET, and age and gender were associated with incident PD.
**We excluded incident PD cases (n = 30) and incident non-PD parkinsonism cases (n = 16).
††Analysis adjusting for age and educational level. These covariates were chosen because in univariate analyses, age and educational level were associated with baseline ET.
‡‡We excluded 30 incident PD cases and the 22 who developed drug induced parkinsonism.
§§One of the two ET cases was a patient with vascular parkinsonism and the other had unspecified parkinsonism.
¶¶Analysis adjusting for age and educational level. These covariates were chosen because in univariate analyses, age and educational level were associated with baseline ET.
ET, essential tremor; PD, Parkinson’s disease.