Table 3.
Risk of incident major CHD or stroke according to usual yogurt intake
| Risk of incident major CHD or stroke | ||||||||
|---|---|---|---|---|---|---|---|---|
| NHS | HPFS | |||||||
| Yogurt intakea | Cases | IR | Age-adjusted HR (95% CI) | Multivariable HR (95% CI)b | Cases | IR | Age-adjusted HR (95% CI) | Multivariable HR (95% CI)b |
| <1/month | 1,793 | 568 | 1.00 | 1.00 | 1,026 | 928 | 1.00 | 1.00 |
| 1/month–<1/week | 661 | 452 | 0.84 (0.77–0.92) | 0.91 (0.83–0.99) | 279 | 733 | 0.89 (0.77–1.01) | 0.92 (0.80–1.05) |
| 1–<2/week | 542 | 429 | 0.77 (0.70–0.85) | 0.85 (0.77–0.94) | 170 | 675 | 0.79 (0.67–0.93) | 0.82 (0.69–0.96) |
| ≥2/week | 479 | 392 | 0.74 (0.67–0.82) | 0.83 (0.74–0.92) | 135 | 679 | 0.77 (0.64–0.92) | 0.79 (0.66–0.96) |
| P for linear trendc | <0.001 | <0.001 | <0.01 | <0.01 | ||||
| Meta-analysis | ||||||||
| <1/month | 2,819 | 662 | 1.00 | 1.00 | ||||
| 1/month–<1/week | 940 | 510 | 0.85 (0.79–0.92) | 0.91 (0.84–0.98) | ||||
| 1–<2/week | 712 | 470 | 0.77 (0.71–0.84) | 0.84 (0.77–0.91) | ||||
| ≥2/week | 614 | 432 | 0.75 (0.68–0.82) | 0.82 (0.75–0.90) | ||||
| P for linear trendc | <0.001 | <0.001 | ||||||
| I 2 | 0.0 | 0.0 | ||||||
Abbreviations: BMI, body mass index; BP, blood pressure; CHD, coronary heart disease; CI, confidence interval; CVD, cardiovascular disease; HBP, high blood pressure; HPFS, Health Professionals Follow-Up Study; HR, hazard ratio; IR, incidence rate per 100,000 person-years; MI, myocardial infarction; NHS, Nurses’ Health Study.
aCumulative average of yogurt intake was calculated from first report of HBP up to the first of the following events: CVD diagnosis, lost to follow up, death, end of study.
bAdjusted for age, race, smoking (defined as never, past, current with 1–14 cigs/day, current with 15–24 c/day, 25+ c/day), family history of MI, physical activity (continuous from baseline), BMI, BP-lowering medication use, and intakes of total energy, alcohol, trans fatty acids, fiber, milk, and cheese.
cLinear trend across yogurt intake categories was quantified with a Wald test for linear trend by assigning the median value to each category and modeling it as a continuous variable.