Table 3.
Cases | Model 1a | Model 1b | Model 2 | ||||
---|---|---|---|---|---|---|---|
n | HR | 95 % CI | HR | 95 % CI | HR | 95 % CI | |
Danish Dietary Guidelines Index | |||||||
Score <3 | 570 | 1·00 | Ref. | 1·00 | Ref. | 1·00 | Ref. |
Score 3–<4 | 1019 | 0·77 | 0·69, 0·85 | 0·87 | 0·78, 0·96 | 0·85 | 0·77, 0·95 |
Score 4–<5 | 484 | 0·66 | 0·58, 0·74 | 0·82 | 0·72, 0·93 | 0·80 | 0·70, 0·90 |
Score ≥5 | 71 | 0·67 | 0·53, 0·86 | 0·91 | 0·71, 1·17 | 0·86 | 0·67, 1·11 |
P for trend* | <0·001 | 0·009 | 0·002 |
Ref., referent category.
All analyses were stratified by date of enrolment (tertiles).
Model 1a: adjusted for length of follow-up (i.e. baseline age).
Model 1b: model 1a plus alcohol intake, alcohol abstinence, physical activity, smoking and education.
Model 2: model 1b plus baseline BMI, waist circumference, hypertension, hypercholesterolaemia and diabetes.
Test for trend was performed by including the categorised index as a linear variable in the model.