Table 3.
Summary of statistically significant race-diet interactions in polytomous logistic regression models estimating the associations of food groups with reproductive status1,2
| AHEI-2010 Subcomponent Score | All Women3 | Black | White | |||
|---|---|---|---|---|---|---|
| βinteraction (SE) | Pinteraction | Odds Ratio | 90% CI | Odds Ratio | 90% CI | |
| Vegetables | ||||||
| PCOS | −0.37 (0.20) | 0.07 | 0.69 | 0.46, 1.03 | 1.45 | 0.98, 2.17 |
| Whole Grains | ||||||
| PCOS | 0.29 (0.16) | 0.07 | 1.34 | 0.98, 1.84 | 0.75 | 0.54, 1.02 |
| OA | 0.22 (0.12) | 0.07 | 1.25 | 0.98, 1.58 | 0.80 | 0.63, 1.02 |
| Sugar-sweetened Beverages, Fruit Juice | ||||||
| OA | 0.15 (0.09) | 0.09 | 1.17 | 0.97, 1.40 | 0.86 | 0.72, 1.03 |
1 AHEI-2010, Alternative Healthy Eating Index 2010; OA Isolated Oligomenorrhea (≥34 days in menstrual cycle), PCOS Polycystic Ovary Syndrome (hyperandrogenism and oligomenorrhea)
2 Model adjusted for covariates: age, total energy intake, education, BMI. Interaction term: diet*race
3 p < 0.10 (each group vs. reference group). Reference group defined as no PCOS, HA, or OA. Baseline group = White. Black sample sizes: PCOS N = 11, OA N = 26, Reference N = 256; race-specific results are derived from the interaction model. White sample sizes: PCOS N = 29, OA N = 49, Reference N = 273; race-specific results are derived from the interaction model