Pooled risk estimates for Parkinson’s disease by exposure assessment method, study design, publication year period and geographic region, based on meta-analysis of articles on occupational pesticide exposure published between 1990 and 2020.
Number of risk estimates | sRR | 95% CI | Heterogeneity measures | ||||
I2 (%) | P value | Q | P value | ||||
Exposure assessment method | 1.20 | 0.55 | |||||
Group-level | 6 | 1.34 | 1.16 to 1.54 | 0 | 0.54 | ||
Self-reported exposure | 22 | 1.45 | 1.18 to 1.76 | 56 | <0.01 | ||
Expert level | 9 | 1.56 | 1.21 to 2.01 | 18 | 0.28 | ||
Study design | 2.82 | 0.24 | |||||
Cohort (prospective) | 8 | 1.28 | 0.95 to 1.73 | 63 | <0.01 | ||
Cohort (retrospective) | 1 | 1.14 | 0.77 to 1.68 | . | . | ||
Case–control | 28 | 1.54 | 1.34 to 1.77 | 27 | 0.09 | ||
Publication year period | 1.49 | 0.22 | |||||
1990–2006 | 19 | 1.58 | 1.32 to 1.89 | 36 | 0.06 | ||
2007–2020 | 18 | 1.34 | 1.12 to 1.62 | 48 | 0.01 | ||
Geographic region | 1.92 | 0.38 | |||||
Europe | 14 | 1.47 | 1.21 to 1.79 | 37 | 0.08 | ||
USA | 19 | 1.53 | 1.24 to 1.88 | 51 | <0.01 | ||
Other | 4 | 1.17 | 0.85 to 1.62 | 30 | 0.23 | ||
Case–control studies only | |||||||
Exposure assessment method | 0.60 | 0.74 | |||||
Group-level | 3 | 1.48 | 1.02 to 2.15 | 0 | 0.39 | ||
Self-reported exposure | 20 | 1.51 | 1.24 to 1.83 | 42 | 0.02 | ||
Expert level | 5 | 1.70 | 1.31 to 2.2 | 0 | 0.88 | ||
Exposure assessment method during publication year periods | |||||||
1990–2006 | |||||||
Exposure assessment method | 1.24 | 0.54 | |||||
Group-level | 3 | 1.57 | 1.09 to 2.27 | 37 | 0.20 | ||
Self-reported exposure | 13 | 1.52 | 1.2 to 1.93 | 37 | 0.09 | ||
Expert level | 3 | 2.38 | 1.12 to 5.03 | 28 | 0.25 | ||
2007–2020 | |||||||
Exposure assessment method | 0.62 | 0.73 | |||||
Group-level | 3 | 1.23 | 0.93 to 1.62 | 0 | 0.83 | ||
Self-reported exposure | 9 | 1.34 | 0.95 to 1.88 | 70 | >0.01 | ||
Expert level | 6 | 1.42 | 1.12 to 1.81 | 0 | 0.49 |
I2=percentage of variation across studies due to heterogeneity. Q=Cochran’s Q.
sRR, summary risk ratio.