Table 6.
Stratified analyses of the association of plasma n-3 and n-6 polyunsaturated fatty acids levels with depression in American adults: NHANES, 2011–2012.a
| PUFA, μmol/L, N | Male | OR (95% CI) | p trendb | Female | OR (95% CI) | p trendb |
|---|---|---|---|---|---|---|
| AA(20:4n-6) | n = 1,039 | n = 1,002 | ||||
| [187.00, 711.80](n = 290) | Ref | 0.68 | [295.00, 718.63](n = 243) | Ref | 0.99 | |
| [711.80, 856.29](n = 273) | 0.92(0.37, 2.29) | [718.63, 878.13](n = 242) | 1.99(1.00, 3.93) | |||
| [856.29, 1033.88](n = 242) | 1.26(0.48, 3.26) | [878.13, 1053.00](n = 250) | 2.44(1.36, 4.39) | |||
| [1033.88, 2570.00](n = 234) | 0.84(0.34, 2.09) | [1053.00, 2560.00](n = 267) | 1.41(0.65, 3.05) | |||
| DHA(22:6n-3) | n = 1,033 | n = 999 | ||||
| [34.30, 99.89](n = 273) | Ref | 0.30 | [25.70, 116.76](n = 176) | Ref | 0.005* | |
| [99.89, 136.00](n = 253) | 2.06(0.53, 8.08) | [116.76, 152.40](n = 227) | 2.24(0.86, 5.88) | |||
| [136.00, 187.07](n = 258) | 1.11(0.25, 4.90) | [152.40, 205.87](n = 261) | 0.50(0.28, 0.87) | |||
| [187.07, 918.00](n = 249) | 1.96(0.91, 4.22) | [205.87, 799.00](n = 335) | 0.49(0.25, 0.96) | |||
| DPA3 (22:5n-3) | n = 1,014 | n = 984 | ||||
| [12.6 40.79](n = 271) | Ref | 0.15 | [15.10, 39.97](n = 291) | Ref | 0.13 | |
| [40.79, 49.90](n = 259) | 1.66(0.49, 5.62) | [39.97, 49.39](n = 238) | 2.06(1.00, 4.25) | |||
| [49.90, 61.18](n = 240) | 2.32(0.78, 6.86) | [49.39, 62.65](n = 211) | 2.08(0.96, 4.51) | |||
| [61.18, 229.00](n = 244) | 2.40(0.78, 7.42) | [62.65, 198.00](n = 244) | 2.39(1.28, 4.47) | |||
| AA/DHA | n = 1,033 | n = 999 | ||||
| [0.92, 4.82](n = 266) | Ref | 0.17 | [0.77, 4.36](n = 309) | Ref | 0.004* | |
| [4.82, 6.52](n = 254) | 2.38(0.81, 7.00) | [4.36, 5.98](n = 278) | 1.94(0.77, 4.88) | |||
| [6.52, 8.12](n = 244) | 2.07(0.73, 5.84) | [5.98, 7.28](n = 243) | 2.32(1.12, 4.81) | |||
| [8.12, 18.91](n = 269) | 0.89(0.19, 4.19) | [7.28, 17.40](n = 169) | 4.09(2.14, 7.84) |
*p < 0.05.
Analyses were adjusted for age, sex, education, race/ethnicity, ratio of family income to poverty and CVH except the stratification variable, and only the statistically significant results were shown after controlling all potential confounders.
Test for trend based on the variable containing a median value for each quartile.