Table 4 Relative risks, quality (Global Quality Index, see explanation in the text) and likelihood for the presence of biases in epidemiological studies on the relationship between Alzheimer's disease and selected occupational exposures.
| Authors/Design/Year/ Place* | Studied exposures. RR for exposure classified as ever/never exposed† | Global quality index | Selection bias‡ | Disease misclassification bias‡ | Exposure misclassification bias‡ | Bias due to surrogates informants‡ | Misclassification of confounding bias ‡ |
|---|---|---|---|---|---|---|---|
| Savitz et al. Retrospective cohort. 1998. USA33 | EMF: aRR = 2.1 (95% CI 0.6 to 6.8) | 62.9 | – – | +/− | – | 9 | – |
| Kukull et al. Case–control. 1995. USA‐Seattle27 | Solvents: Men: aRR = 6.3 (95% CI 2.2 to 18.1). Women: aRR = 0.6 (95% CI 0.2 to 1.9). Both: aRR = 1.8 (95% CI 1.1 to 3.1) | 55.6 | – | – | – | +/− | – |
| Tyas et al. Prospective cohort. 2001. Manitoba, Canada38 | Solvents: Degreasers: aRR = 0.88 (95% CI 0.31 to 2.50) Pesticides: Defoliants, fumigants: aRR = 4.35 (95% CI 1.05 to 17.90). Pesticides/fertilisers: aRR = 1.45 (95% CI 0.57 to 3.68) | 53.8 | +/− | – | + (?) | 9 | – |
| Baldi et al. Prospective cohort. 2003. Paquid study‐France40 | Pesticides: Men: aRR = 2.39 (95% CI 1.02 to 5.63). Women: aRR = 0.89 (95% CI 0.49 to 1.62) | 50.5 | +/− | – | +/− | 9 | +/− |
| Salib and Hillier. Case–control. 1996. UK28 | Aluminium: Demented controls: aRR = 0.95 (95% CI 0.5 to 1.8). Non‐demented controls: aRR = 0.95 (95% CI 0.5 to 1.9) | 48.1 | – – | +/− | + (↓) | +/− | – |
| Savitz et al. Case–control. 1998. USA‐25 different states35 | EMF: Electrical/non‐electrical occupation: aRR = 1.2 (95% CI 1.0 to 1.4) | 44.4 | – | + (↓) | ++ (↓) | 99 | +/− |
| Graves et al.Case–control. 1999. USA‐Seattle36 | EMF: Hygienist 1: aRR = 0.74 (95% CI 0.29 to 1.92); hygienist 2: aRR = 0.95 (95% CI 0.29 to 1.92) | 42.6 | +/− | – | +/− | + (↑) | +/− |
| Chandra et al. Case–control. 1987. USA‐Denver13 | Lead: aRR = 0.25 (95% CI 0.03 to 2.24)§ | 41.7 | +/− | +/− | + (↓) | + (?) | + (?) |
| Gauthier et al. Case–control. 2001. Canada‐Quebec37 | Pesticides: aRR = 0.97 (95% CI 0.38 to 2.41). Herbicides: aRR = 1.07 (95% CI 0.39 to 2.54). Insecticides: aRR = 1.62 (95% CI 0.64 to 4.11) | 41.7 | +/− | +/− | +/− | +/− | – |
| CSHA** et al. Case–control. 1994. 10 Canadian provinces26 | Solvents: aRR = 0.76 (95% CI 0.38 to 1.54) Pesticides: Pesticides/fertilisers: aRR = 1.58 (95% CI 0.81 to 3.10) | 39.8 | +/− | – | + (↓) | + (?) | +/− |
| O'Flynn et al. Case–control. 1987. England and Wales24 | Solvents: unadjusted RR = 1.11 (not significant) Lead: unadjusted RR = 0.86 (not significant) | 39.8 | – | ++ (↓) | ++ (↓) | 99 | + (?) |
| Graves et al. Case–control. 1998. USA‐Seattle32 | Solvents: aRR = 1.77 (95% CI 0.81 to 3.90) Aluminium: unadjusted RR = 1.46 (95% CI 0.63 to 3.42) | 37.0 | +/− | – | + (↓) | +/− | + (?) |
| Feychting et al. Case–control. 1998. Sweden34 | EMF: Control group 1:aRR = 2.4 (95% CI 0.8 to 6.9). Control group 2: aRR = 2.7 (95% CI 0.9 to 7.8). | 36.1 | +/− | – | +/− | + (?) | +/− |
| Noonan et al. Case–control. 2002. USA‐Colorado39 | EMF: aRR = 1.21 (95% CI 0.83 to 1.76) | 34.3 | +/− | ++ (↓) | ++ (↓) | 99 | +/− |
| Sobel et al. 1996. Case–control. USA‐California29 | EMF: Men: aRR = 4.90 (95% CI 1.3 to 7.9). Women: aRR = 3.40 (95% CI 0.8 to 16). Both: aRR = 3.93 (95% CI 1.45 to 10.56) | 34.2 | + (↓) | +/− | ++ (↓) | 99 | + (?) |
| Shalat et al. 1988. Case–control. USA‐Bedford14 | Solvents: aRR = 1.0 (95% CI 0.5 to 1.9) Lead: aRR = 0.8 (95% CI 0.3 to 2.0) | 33.3 | + (?) | +(?) | + (?) | ++ (?) | +/− |
| Li et al. Case–control. 1992. China25 | Solvents: aRR = 1.17 (95% CI 0.31 to 4.37) | 30.6 | + (?) | +/− | + (↓) | ++ (↓) | + (?) |
| French et al. Case–control. 1985. USA‐Minneapolis12 | Solvents: aRR = 1.25 (95% CI 0.55 to 2.84) Pesticides: aRR = 0.80 (95% CI 0.29 to 2.19) Lead: aRR = 1.50 (95% CI 0.25 to 8.98)** | 28.7 | + (?) | + (?) | + (↓) | + (?) | ++ (?) |
| Gun et al. Case–control. 1997. Australia30 | Solvents: unadjusted RR = 1.31 (95% CI 0.83 to 2.07 Pesticides: Organophosphates: unadjusted RR = 2.54 (95% CI 0.41 to 27.06) Lead: unadjusted RR = 1.12 (95% CI 0.63 to 2.0) Aluminium: unadjusted RR = 0.33 (95% CI 0.01 to 4.16) | 26.8 | ++ (?) | + (?) | + (↓) | ++ (↓) | +/− |
| Heyman et al. Case–control. 1984. USA‐Durhan11 | Solvents: 0 exposed¶ Lead: aRR = 0.78 (95% CI 0.14 to 4.36)¶ | 24.1 | ++ (?) | + (↓) | ++ (↓) | +/− | +/− |
| Series 2. Sobel et al15 | EMF (see series 1 data) | 24.1 | ++ (?) | + (?) | + (↓) | ++ (↓) | +/− |
| Series 3. Sobel et al15 | EMF (see series 1 data) | 21.3 | ++ (?) | + (?) | + (↓) | ++ (↓) | +/− |
| Series 1.Sobel et al. Case–control. 1995. Finland and USA‐California15 | EMF: For the three series: Men: aRR = 1.9 (95% CI 0.8 to 5.0). Women: aRR = 3.7 (95% CI 1.7 to 8.9). Both sexes: aRR = 2.9 (95% CI 1.6 to 5.4) | 20.4 | ++ (?) | + (?) | + (↓) | ++ (↓) | +/− |
| Palmer et al. Case–control. 1998. England and Wales31 | Solvents: aRR = 0.3 (95% CI 0.1 to 1.3) | 19.4 | + (↓) | ++ (↓) | ++ (↓) | ++ (↓) | + (?) |
*Papers are listed in decreasing order of their Global Quality Index, see explanation in the text.
†Highest relative risks (and odds ratios) adjusted by the maximum number of variables in each study.
‡For bias: – – denotes highly improbable, – denotes improbable, +/− denotes possible, + denotes probable, ++ denotes highly probable. For each bias identified as probable or highly probable, (↓) denotes the experts either judged that the bias would had decreased association towards the null, (↑) denotes bias would increase the association, and (?) denotes the experts failed to reach a conclusion about the effect of that particular bias. 9: Registry‐based retrospective or prospective cohorts in which this bias is not possible; 99: Registry‐based case–control studies, in which this bias is not possible.
§Information about OR for lead was obtained from a later reanalysis.10
¶Information about OR for solvents and lead was obtained from a later reanalysis.10
**CSHA, Canadian Study of Health and Aging Investigators.