You’re quoting from the discussion in this paper. This passage cites five references. Despite what the passage says, none of them is a meta-analysis.
[40] is a small RCT (33 people) from Korea, and did find that fish oil lowered apoB-100 by 24%. This is an exception rather than the rule on the claims of the authors of the cited passage.
[41] Was another small RCT, this one in men with hypertriglyceridemia (average TG 225.66 mg/dL): “Thirty-four men (17 in each group) completed the study, but for the analyses reported here, we had samples from only 14 subjects in each group.” The cited paper does not actually contain the effects on lipids, but refers back to their previous paper, which found that DHA “elevated concentrations of LDL cholesterol (12.6%), small VLDL particles (133%), and large LDL particles (120%) and the mean diameter of LDL particles (0.6 nm) in fasting plasma.” So the exact opposite of what the paper from which you’re quoting says.
[42] is a review on the mechanisms whereby omega-3s affect lipid metabolism: its only section on levels says “Comparing EPA treatment with DHA treatment, or DHA treatment with OM3FA treatment, generally shows similar and small or no effects on LDL-C levels [18, 19, 23,24,25,26]. However, it is reported that treatment with DHA, but not EPA, increases LDL-C levels and LDL particle size [22, 54]. In contrast, one study showed no effect of DHA, but EPA treatment decreased TG levels and increased LDL-C levels in normolipidemic individuals [20].” So certainly no support for it lowering LDL-C/apob.
[43] was an RCT that “randomly assigned [metabolic syndrome] patients to 1 of 4 diets: high-SFA [HSFA; 38% energy (E) from fat, 16% E as SFA], high-monounsaturated fatty acid [HMUFA; 38% E from fat, 20% E as MUFA], and 2 low-fat, high-complex carbohydrate [LFHCC; 28% E from fat] diets supplemented with 1.24 g/d of long-chain (LC) (n-3) PUFA (ratio 1.4 eicosapentaenoic acid:1 docosahexaenoic acid) or placebo (1.24 g/d of high-oleic sunflower-seed oil) for 12 wk each.” At the end of the trial, the LFHCC with fish oil led to a nonsignificant −5.3% reduction in postprandial LDL-C and a nonsignificant 5.9% increase in postprandial apoB48; no data are given on total apoB, apoB-100, or fasting lipid values.
[44] is a review about omega-3 fatty acids in CVD; the only thing it says about lipoproteins is that “In patients with this degree of TG elevation [135 to 499 mg/dl], a 20 to 30% reduction in fasting [TG] values is typical and often accompanied by small reductions in nonhigh-density lipoprotein cholesterol and apolipoprotein B.”
So the quoted passage misrepresents or exaggerates its cited studies.
As I noted, none of the cited studies is a meta-analysis. This study actually was a meta-analysis. It forund that omega-3 fatty acids had no effect on LDL-C (WMD 0.03 [—0.08 to 0.15]).
Reviewing this I’ll retract my statement that FO elevates apoB/LDL-C: it appears that it has no effect. But it certainly doesn’t meaningfully lower it.