Table 2.
NutriGrade assessment of confidence in estimate effect of studies evaluated the association between various food groups and risk of BC.
| Food groups | Risk of bias 1 | Precision 2 | Indirectness | Heterogeneity | Publication bias 3 | Effect size | Dose response | Funding bias | Total score | Confidence evidence 4 |
|---|---|---|---|---|---|---|---|---|---|---|
| Fruits | 2 | 0 | 1 | 1 | 1 | 0 | 0 | 1 | 6 | Moderate |
| Vegetables | 2 | 0 | 1 | 1 | 1 | 0 | 0 | 1 | 6 | Moderate |
| Legumes | 2 | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 4 | Low |
| Egg | 2 | 0 | 1 | 0 | 0 | 0 | 1 | 1 | 5 | Low |
| Dairy | 2 | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 4 | Low |
| Fish | 2 | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 4 | Low |
| Red meat | 2 | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 4 | Low |
| Processed meat | 2 | 0 | 1 | 0.5 | 0.5 | 0 | 0 | 1 | 5 | Low |
| Sugar-sweetened drinks | 2 | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 4 | Low |
| Alcohol | 2 | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 4 | Low |
| Tea | 2 | 0 | 1 | 1 | 1 | 0 | 1 | 1 | 7 | Moderate |
| Coffee | 2 | 0 | 1 | 1 | 1 | 0 | 0 | 1 | 6 | Moderate |
NutriGrade, Nutrition Grading of Recommendations Assessment, Development and Evaluation; RR, risk ratio.
Risk of bias was based on the Newcastle-Ottawa Scale, where ≥7 = 2 points; 4–6.9 = 1 point; and 0–3.9 = 0 points.
Precision is 1 point if the number of events ≥500 and the 95% CI excludes the null value; precision is 0 points if the number of events <500 or number of events ≥500, but 95% CI includes the null value and 95% CI fails to exclude an important benefit (RR of 0.8) or harm (RR of 1.2).
Based on the funnel plots, Egger or Begg’s test. For the outcomes with small number of studies (n < 10), the risk of publication bias was not formally assessed.
High quality indicates that there is high confidence in the effect estimate, and further research probably will not change the confidence in the effect estimate. Moderate quality indicates that we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low quality indicates that our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.