Table 2.
Main characteristics and results (for the whole study group, or stratified by menopausal status when available) of cohort studies reporting on the association between a-priori- or a-posteriori-defined dietary patterns and mammographic breast density.
| Author, Year | Study Name and Country | Study Population | Dietary Assessment | Dietary Pattern Type | Dietary Pattern | Summary of Results (a) |
|---|---|---|---|---|---|---|
| Garzia et al., 2021 [25] | NHS-II, USA | Pre-menopausal (n = 1117, median age 44 y) | Adult-FFQ (diet in early adulthood, filled at 27–44 y) | A priori | Pro-inflammatory pattern | Mean MBD 40.7% (38.6–42.8) vs. 40.6% (38.6–42.7) for the 5th vs. 1st quintile, p-value for trend 0.99 |
| AHEI anti-inflammatory pattern | Mean MBD 39.9% (37.9–42.0%) vs. 39.7% (37.7–41.7%) for the 5th vs. 1st quintile, p-value for trend 0.98 | |||||
| Pre-menopausal (n = 709, median age 44 y) | HS-FFQ (diet during high school, filled at 33–52 y) | Pro-inflammatory pattern | Mean MBD 38.7% (35.7–41.7%) vs. 39.6% (37.3–41.9%) for the 5th vs. 1st quintile, p-value for trend 0.93 | |||
| AHEI anti-inflammatory pattern | Mean MBD 41.4% (38.8–43.9%) vs. 41.0% (38.3–43.6%) for the 5th vs. 1st quintile, p-value for trend 0.89 | |||||
| Mishra et al., 2011 [26] | The Medical Research Council National Survey of Health and Development, UK | Pre- and post-menopausal (n = 792, mean age 51.5, SD = 1.1 y) | Maternal 24 h recalls (about diet at 4 y) | A posteriori | breads and fats | β −0.004 (−0.08, 0.07) |
| fried potatoes and fish | β −0.05 (−0.12, 0.01) | |||||
| milk, fruit and biscuits | β −0.01 (−0.08, 0.05) | |||||
| Pre- and post-menopausal (n = 700, mean age 51.5, SD = 1.1 y) | 5-day food diaries (about diet at 36 and 43 y) | low fat, high fiber | β 0.03 (−0.04, 0.11) | |||
| alcohol and fish | β −0.02 (−0.13, 0.17) | |||||
| high fat and sugar | β 0.06 (−0.01, 0.13) | |||||
| meat, potatoes, and vegetables | β −0.03 (−0.10, 0.04) |
(a) Most adjusted results odds ratio (OR, for logistic regression models), or â coefficients (for linear regression models), along with corresponding 95% confidence intervals (CI) and/or p-values, were reported whenever available. Lacking this information, we reported the mean/median MBD in the groups being compared, along with the p-values for comparison (if available). AHEI, Alternative Healthy Eating Index; FFQ, food-frequency questionnaire; MBD, mammographic breast density; NHS-II, Nurses’ Health Study II; SD, standard deviation.