Fish vs. Chicken vs. Plant Protein for Fatty Liver: A Data Comparison

'What should I eat for protein?' is the question that follows every fatty liver diagnosis, and the honest answer is not a single food — it's a hierarchy informed by three data points: saturated fat load, omega-3 content, and processing level. Let me rank the protein sources the way the evidence does, then get into the trade-offs.

At the top of the hierarchy sits fish, specifically fatty fish — salmon, sardines, mackerel, herring. The case is unusually strong. Fatty fish delivers high-quality protein with saturated fat under 25% of total fat (a 100g salmon serving has roughly 3g of saturated fat against 13g of total fat), plus omega-3 fatty acids — EPA and DHA — that have direct anti-inflammatory effects on the liver and appear to lower liver triglyceride levels in intervention studies. The AASLD and Mediterranean framework both place fish at the top, with the consistent recommendation of two to three servings per week. Sardines deserve special mention: they're among the most omega-3-dense foods per gram, cheap, and low in the mercury concern that applies to some larger fish.

Second is poultry — skinless chicken and turkey. The data is straightforward: poultry is lean, low in saturated fat (chicken breast sits around 25% saturated as a proportion of fat, with very low absolute fat), and neutral on the omega-3 axis — it neither helps nor hurts on that front. It's the workhorse protein: easy, cheap, familiar, and reliably moderate. The caveats are processing and cooking: chicken skin adds saturated fat, fried chicken adds oil and often breading (refined carbohydrate), and processed chicken products (nuggets, sausages, deli slices) import the processed-meat problems. Whole, skinless, baked or grilled chicken is the version the data supports.

Third, plant proteins — legumes, tofu, tempeh, and edamame. This is where the verdict gets interesting, because plant proteins are structurally different rather than merely lower: they deliver protein wrapped in fiber, with essentially no saturated fat (soy has a few grams, all unsaturated) and zero cholesterol. The fiber is the hidden asset — it feeds the gut microbiome, produces anti-inflammatory short-chain fatty acids, and slows the absorption of anything eaten alongside. Legumes like lentils and chickpeas are also the only protein source that simultaneously counts as a whole grain substitute, which makes them doubly valuable in the Mediterranean framework. The downside is partial protein quality — most plant proteins are lower in some essential amino acids than animal proteins — but that's only relevant if plants are your sole protein source, which few people's diets are.

Now the comparisons people actually ask about. Eggs: fine, roughly equivalent to poultry on the saturated fat axis (about 1.6g per large egg) with excellent protein quality; the old cholesterol panic doesn't survive contact with the liver data. Dairy protein: yogurt and cottage cheese are good; cheese and butter are saturated-fat concentrated and belong at the low end. Red meat: the most nuanced, covered in detail in our red meat guide — fine as lean cuts, twice a week, with processed versions effectively excluded. Protein powders and bars: the wild card — a good whey or plant protein shake can slot in as a convenience protein with zero saturated fat, but many bars import palm oil, chocolate, and added sugar, so the label method applies.

The synthesis the evidence supports is a rotation rather than a single answer: fish two to three times a week for the omega-3 benefit, poultry and eggs as the everyday workhorses, legumes and tofu several times a week for the fiber bonus, and red meat as the occasional lean choice. That pattern is basically the Mediterranean protein architecture, which is exactly why the guidelines point there — it's not that any single protein is magic, it's that the combination covers every axis at once.

One honest caveat before the bottom line: 'plant-based' is not automatically liver-friendly. The category includes tofu and lentils — excellent — but also plant-based burgers that are saturated-fat bombs in disguise, and plant-based 'cheese' that's largely coconut oil (over 80% saturated). The plant label is not a liver-health label; the nutrient profile is. As with everything else in this disease, the data follows the food, not the marketing.

A note on cost and accessibility, because protein advice that ignores price is advice that fails in practice. The hierarchy above has an affordable expression: canned sardines and mackerel deliver the omega-3 benefit at a fraction of fresh salmon's price; eggs remain the cheapest high-quality protein on the market; lentils, chickpeas, and tofu are among the least expensive calories-per-gram foods in any grocery store; and frozen chicken breast and frozen fish fillets cost meaningfully less than fresh while retaining the nutrient profile. The Mediterranean framework is sometimes framed as an expensive diet — it doesn't have to be. The data supports the pattern, and the pattern has a budget version that tracks the same nutrient priorities.

Finally, the portion question, because it's where otherwise-correct protein choices go sideways. The serving framework that shows up across sources — Mayo Clinic, NHS, and the Mediterranean references — runs roughly 85-150g of cooked protein per meal depending on the source. A serving is about the size of a palm or a deck of cards; the 'steakhouse cut' of 250-300g is two to three servings in one sitting, which quietly defeats the saturated-fat discipline of the whole pattern. Portion discipline is not an extra rule — it's how the hierarchy survives contact with real meals. A moderate serving of the right protein, on a plate built around vegetables, is the shape of every data-supported meal in this disease.

I've fielded the what-should-I-eat-for-protein question from enough people to know the real problem isn't knowledge — it's rotation. People pick one protein and eat it every day, then get bored and quit. The data supports a rotation precisely because it keeps you engaged. Sustainability beats optimization every time.

Here's the takeaway I'd want if I were reading this: stop looking for the single best protein and start rotating. Fish two to three times a week for the omega-3s, poultry and eggs as everyday workhorses, legumes and tofu for the fiber bonus, and lean red meat as the occasional choice. The nutrient data — not the marketing label — decides each slot. And the affordable version of this pattern exists: canned sardines, eggs, lentils, and frozen fish track the same priorities at a fraction of the price.

Let me give you the shopping-list version of this hierarchy, because that's where decisions actually happen. Frozen salmon and sardines are the omega-3 workhorses that fit a budget. Chicken thighs are fine if the skin is removed after cooking — the fat lives in the skin, not the meat. Lentils and chickpeas are the cheapest protein in the store and double as the fiber source the guidelines want. Eggs are the protein that survives any budget. The pattern isn't expensive; the pattern is deliberate.

One clarification that matters: the omega-3 benefit is specifically from EPA and DHA, the marine forms found in fish. Plant sources like flax and chia provide ALA, which the body converts to EPA/DHA only inefficiently — roughly 5-10% conversion. So a plant-based person doesn't get the same omega-3 liver signal from flax as someone eating sardines. That's not a reason to eat fish if you're plant-based; it's a reason to know that the plant-based omega-3 case is different, and the fiber and saturated-fat benefits of plants carry the argument instead.

I've fielded the 'what should I eat for protein' question from enough people to know the real problem isn't knowledge — it's rotation. People pick one protein and eat it every day, then get bored and quit. The data supports a rotation precisely because it keeps you engaged. Sustainability beats optimization every time.

How we compared the proteins

Protein comparisons are one of the most-read topics on the site, so we built this with the rating engine as the referee. Our data editor pulled the fish, poultry, and plant protein entries, compared saturated shares and processing levels, and the ranking emerged from the numbers: most fish and skinless poultry in Recommended, red meat cuts varying by trim, processed meats in Limit. The plant proteins (legumes especially) rated well on fiber and saturated share.

We also checked the omega-3 angle for fish because it adds a dimension the saturated-share rule does not charge for. The comparison is data-driven, and we say which criteria decided each entry.

Frequently Asked Questions