Table 3.
Association of Log Omega-3 Fatty Acid Levels (as a Continuous Variable per 1-SD Increase) With Leukocyte Telomere Shortening, Defined as >10% Decrease in Telomere Length Over 5 Years
| Modela | Odds Ratio per 1-SD Increase (95% Confidence Interval) | P Value |
|---|---|---|
| Unadjusted | 0.81 (0.69-0.95) | .01 |
| Model 1 | 0.72 (0.60-0.88) | .001 |
| Model 2 | 0.68 (0.54-0.85) | .001 |
| Model 3 | 0.68 (0.54-0.86) | .001 |
| Model 4 | 0.68 (0.54-0.86) | .001 |
| Model 5 | 0.68 (0.47-0.98) | .04 |
All covariates were modeled continuously except the following categorical variables: sex (male, female); ethnicity (Hispanic, Asian, black, white, other); income (<$10 000, $10 000-19 999, $20 000-29 999, $30 000-39 999, $40 000-50 000, ≥$50 000); education (no formal schooling, 5th grade or less, 6th-8th grade, 9th-11th grade, high school graduate, some college/vocational school, college degree, graduate/professional degree); prior myo-cardial infarction; type 2 diabetes; current smoking status; vitamin use; statin use; β-blocker use. In the final adjusted model (model 5), the linear regression equation is log odds of telomere shortening=−1.06−0.87(log omega-3 level)+0.05(age)+0.87(male)+0.68(Asian)+0.91 (black)+0.57(white)+0.58(other)−0.3(income $10 000-$19 999)−0.008(income $20 000-$29 999)+0.56 (income $30 000-$39 999)−0.44(income $40 000-$49 999)−0.14(income ≥$50 000)−16(≤5th grade)−18 (6th-8th grade)−18(9th-11th grade)−18(high school graduate)−18(some college)−18(college degree)−18 (graduate degree)+0.07(exercise capacity)+0.39(prior myocardial infarction)+0.24(type 2 diabetes)+0.09 (smoking)+0.01(left ventricular ejection fraction)+4.0 (waist-hip ratio)+0.01(systolic blood pressure)−0.03 (diastolic blood pressure)−0.05(log triglycerides)+0.54 (log low-density lipoprotein cholesterol)−0.11(log high-density lipoprotein cholesterol)−0.005(log C-reactive protein)+0.09(log interleukin 6)+0.64(multivitamin use) −0.21(statin use)−0.35(β-blocker use)+10(baseline telomere length)+0.0002(baseline telomere length×log omega-3 level). Model 1 was adjusted for age, sex, race/ethnicity, income, education, and exercise capacity. Model 2=model 1 + prior myocardial infarction, type 2 diabetes, smoking, left ventricular ejection fraction, waist-hip ratio, systolic and diastolic blood pressure, log triglycerides, log low-density lipoprotein cholesterol, and log high-density lipoprotein cholesterol. Model 3=model 2 + log C-reactive protein and log interleukin 6. Model 4=model 3 + use of vitamins, statins, and β-blockers. Model 5=model 4 + baseline telomere length and baseline telomere length × log omega-3 level.