Table 3.
Association of seafood n-3 PUFA biomarkers with secondary outcome of ≥40% decrease in eGFR from baseline
| n-3 PUFA biomarker (No of studies; participants with ≥40% decrease in eGFR)* | Model† | Per interquintile range | Highest fifth v lowest fifth | ||||
|---|---|---|---|---|---|---|---|
| Relative risk (95% CI) | I2 (%) | Relative risk (95% CI) | I2 (%) | ||||
| EPA (16; 2554) | 1 | 0.99 (0.91 to 1.07) | 0.0 | 0.85 (0.74 to 0.98) | 19.5 | ||
| 2 | 0.98 (0.91 to 1.06) | 7.6 | 0.84 (0.73 to 0.97) | 26.1 | |||
| DPA (13; 2110)) | 1 | 0.92 (0.82 to 1.02) | 42.5 | 0.85 (0.73 to 0.98) | 21.8 | ||
| 2 | 0.92 (0.82 to 1.03) | 42.1 | 0.86 (0.74 to 1.00) | 33.7 | |||
| DHA (16; 2554) | 1 | 0.93 (0.84 to 1.03) | 52.2 | 0.88 (0.76 to 1.01) | 46.2 | ||
| 2 | 0.93 (0.83 to 1.03) | 54.3 | 0.88 (0.76 to 1.02) | 47.1 | |||
| EPA+DPA+DHA‡ (16; 2552) | 1 | 0.95 (0.86 to 1.04) | 45.2 | 0.85 (0.74 to 0.98) | 44.1 | ||
| 2 | 0.94 (0.85 to 1.03) | 46.4 | 0.85 (0.74 to 0.98) | 45.3 | |||
| ALA (16; 2553) | 1 | 0.98 (0.89 to 1.07) | 35.8 | 0.97 (0.85 to 1.12) | 7.1 | ||
| 2 | 0.94 (0.86 to 1.04) | 18.9 | 0.95 (0.82 to 1.09) | 0.0 | |||
Effect estimates were pooled using inverse variance weighted meta-analysis.
ALA=α linolenic acid; CKD=chronic kidney disease; DHA=docosahexaenoic acid; DPA=docosapentaenoic acid; eGFR=estimated glomerular filtration rate; EPA=eicosapentaenoic acid; n-3 PUFA=omega 3 polyunsaturated fatty acid.
Number of cohorts contributing to this analysis was lower than the primary outcome because three cohorts were excluded due to low number of participants with ≥40% decrease in eGFR (FDPS—Finnish Diabetes Prevention Study, MAS—Memory and Ageing Study, and METSIM—Metabolic Syndrome in Men). The small difference in number of participants with ≥40% decrease in eGFR was due to missing measurement for specific n-3 PUFA fatty acids in some cohorts.
Model 1 adjusted for age, sex, race, clinical centre or field site, education, occupation, body mass index, smoking, alcohol intake, physical activity, prevalent coronary heart disease, and use of lipid lowering drugs, when applicable. Model 2 adjusted for all covariates in model 1 and also adjusted for prevalent diabetes mellitus, urine albumin-creatinine ratio, systolic blood pressure, use of angiotensin converting enzyme inhibitors or angiotensin receptor blockers, and use of other antihypertensive drugs.
DPA was not available in three of the cohorts (CCCC—Chin-Shan Community Cardiovascular Cohort, InCHIANTI—Invecchiare in Chianti Study, and ULSAM—Uppsala Longitudinal Study of Adult Men), therefore the sum in these cohorts was calculated as EPA+DHA.