TABLE 5.
Effect | 95% CI | SE | |
Direct effect | |||
Dietary long-chain n – 3 PUFAs → breast cancer | 0.282 | –1.261, 1.825 | 0.787 |
Indirect effect | |||
(1) Dietary long-chain n – 3 PUFAs → RALA → breast cancer | –0.060 | –0.237, 0.070 | 0.075 |
(2) Dietary long-chain n – 3 PUFAs → REPA → breast cancer | –0.409 | –0.853, –0.018 | 0.212 |
(3) Dietary long-chain n – 3 PUFAs → RDPA → breast cancer | 0.002 | –0.126, 0.143 | 0.064 |
(4) Dietary long-chain n – 3 PUFAs → RDHA → breast cancer | –0.167 | –0.361, –0.025 | 0.086 |
(5) Dietary long-chain n – 3 PUFAs → RALA → REPA → breast cancer | –0.011 | –0.049, 0.011 | 0.015 |
(6) Dietary long-chain n – 3 PUFAs → RALA → RDPA → breast cancer | –0.004 | –0.018, 0.005 | 0.006 |
(7) Dietary long-chain n – 3 PUFAs → RALA → RDHA → breast cancer | –0.001 | –0.009, 0.005 | 0.003 |
(8) Dietary long-chain n – 3 PUFAs → REPA → RDPA → breast cancer | –0.167 | –0.332, –0.033 | 0.076 |
(9) Dietary long-chain n – 3 PUFAs → REPA → RDHA → breast cancer | –0.139 | –0.237, –0.063 | 0.044 |
(10) Dietary long-chain n – 3 PUFAs → RDPA → RDHA → breast cancer | –0.0002 | –0.015, 0.015 | 0.007 |
(11) Dietary long-chain n – 3 PUFAs → RALA → REPA → RDPA → breast cancer | –0.004 | –0.018, 0.005 | 0.006 |
(12) Dietary long-chain n – 3 PUFAs → RALA → REPA → RDHA → breast cancer | –0.004 | –0.014, 0.004 | 0.005 |
(13) Dietary long-chain n – 3 PUFAs → RALA → RDPA → RDHA → breast cancer | –0.0004 | –0.002, 0.001 | 0.001 |
(14) Dietary long-chain n – 3 PUFAs → REPA → RDPA → RDHA → breast cancer | –0.018 | –0.041, –0.005 | 0.009 |
(15) Dietary long-chain n – 3 PUFAs → RALA → REPA → RDPA → RDHA → breast cancer | –0.001 | –0.002, 0.001 | 0.001 |
Total indirect effect | |||
Dietary long-chain n – 3 PUFAs → breast cancer | –0.982 | –1.529, –0.508 | 0.261 |
RALA, erythrocyte α-linolenic acid; CI, confidence interval; RDHA, erythrocyte docosahexaenoic acid; RDPA, erythrocyte docosapentaenoic acid; REPA, erythrocyte eicosapentaenoic acid; PUFA, polyunsaturated fatty acid; SE, standard error.
aAdjusted for age, BMI, MET-h/week, education, passive smoking, regular drinking, first-degree relatives with cancer and energy intake.