Table 4.
Comparison of relative rates of nuclear sclerotic, cortical and posterior subcapsular cataract among 20,599 participants in the Physicians’ Health Study (US male physicians, 1982–1995)
| Cataract Type |
||||
|---|---|---|---|---|
| Nuclear Sclerotic (1,726 events) | Cortical (734 events) | Posterior Subcapsular (794 events) | Test of equal associations across types* | |
| HR (95% CI) | HR (95% CI) | HR (95% CI) | P-value | |
| BMI, per 1 kg/m2 | 1.03 (1.01–1.05) | 1.00 (0.97–1.03) | 1.06 (1.03–1.08) | 0.002 |
| Multivitamin use | 0.90 (0.80–1.00) | 0.80 (0.67–0.96) | 0.89 (0.75–1.05) | 0.34 |
| Family history of MI | 1.12 (0.96–1.30) | 1.02 (0.80–1.30) | 1.27 (1.04–1.56) | 0.20 |
| Diabetes mellitus | 1.23 (0.96–1.57) | 1.67 (1.20–2.32) | 1.76 (1.29–2.40) | 0.003 |
| Smoke <1 pack/day | 1.41 (1.11–1.80) | 1.26 (0.85–1.87) | 1.56 (1.11–2.19) | 0.71 |
| Smoke ≥1 pack/day | 1.49 (1.23–1.79) | 1.57 (1.18–2.08) | 1.70 (1.32–2.20) | |
| Former smoker | 1.12 (1.01–1.23) | 1.19 (1.02–1.38) | 1.16 (1.00–1.35) | |
Note: model also allows for different effects of randomized treatment assignment, treats age as the time scale and stratifies on cataract type
The null hypothesis is that the risk factor has the same association for all cataract types; for smoking the null hypothesis assumes each smoking variable has a uniform association across types