Karpathy cited a tweet/video about choosing low-emulsifier almond milk (at least whole foods contains SOME, but NOT AS MUCH as some other brandsā¦).
Idk, I dont care about almond milk emulsifiers that much, I need food to be filling but like **stay tf away from primal kitchen avocado salad dressing w/emulsifiers
Do whole eggs ācontain emulsifiersā?
Yes ā egg yolk is itself a powerful, natural emulsifier.
Where the āemulsifierā lives |
Main molecules |
Typical amount in one large egg (ā 50 g) |
What they do |
Yolk |
⢠Phosphatidylcholine (āālecithinā) ⢠Phosphatidylethanolamine |
ā 1.3 g total phospholipids, of which ~ 0.8ā1 g is phosphatidylcholine (ā 115ā170 mg choline) (PubMed Central, PubMed Central) |
Amphiphilic heads + hydrophobic tails sit at the oilāwater interface, lowering surface tension and stabilising the emulsion (e.g. mayonnaise, hollandaise). (American Egg Board, MDPI) |
White |
Proteins such as ovalbumin & ovotransferrin |
ā |
Provide some secondary emulsifying & foaming capacity, but far weaker than yolk. |
So when you crack a whole egg into salad dressing or a blender, youāre supplying the emulsifier straight from the food itselfānothing āaddedā or synthetic.
How this differs from āadded emulsifiersā on labels
Natural egg-yolk lecithin |
Common processed-food emulsifiers (CMC, polysorbate-80, mono- & diglycerides) |
Chemical class: phospholipid exactly like those in human cell membranes. |
Synthetic surfactants or modified cellulose/fats not normally eaten in bulk. |
Dose in a dish: typically ⤠1 % of the finished mayo/dressing. |
Can reach 3ā5 % of product weight in ultra-processed foods. |
Microbiome evidence: no data linking native yolk lecithin to mucus erosion; most studies focus on large supplemental choline ā TMA/TMAO, a different pathway. |
CMC & polysorbate-80 repeatedly thin the mucus layer and drive dysbiosis in mice; epidemiology now probing humans. |
Regulatory status: āwhole eggā or āegg yolkā on the ingredient line (nothing more). |
Must appear individually on the label as emulsifiers. |
In short, eating eggs gives you lecithin in its natural matrix, accompanied by protein, cholesterol, vitamins A/D/E/K, and carotenoids. Thatās a world apartābiochemically and by doseāfrom purified emulsifiers sprinkled into plant-based milks or candy coatings.
Practical take-aways
- Whole eggs already are an emulsifier; nothing additional is hidden inside.
-
āLecithinā on an ingredients list usually means soy- or sunflower-derived phosphatidylcholine extracted and dried. Egg-derived lecithin exists but is costly, so the additive is almost always plant-based.
-
If youāre avoiding microbiome-disrupting emulsifiers, there is no evidence you need to exclude normal culinary use of eggs. The main gut-related caution with eggs concerns high choline ā higher TMAO in some people, not mucus-layer stripping.
Feel free to ask if you want more on choline/TMAO metabolism or how egg proteins contribute to foams and emulsions in baking.
===
Step 1 ā roughly how much emulsifier youāre actually getting
Daily almond-milk habit |
Volume |
Typical gellan-gum level (0.02 ā 0.03 % w/w) |
Sunflower-lecithin level (order-of-magnitude estimate¹) |
Total emulsifier load |
½ gal (1.89 L) |
1 900 g |
0.38 ā 0.57 g gellan |
ā 0.8 ā 1.6 g lecithin |
~1.2 ā 2.2 g |
¾ gal (2.84 L) |
2 840 g |
0.57 ā 0.85 g gellan |
ā 1.2 ā 2.4 g lecithin |
~1.8 ā 3.3 g |
¹Blue Diamond, Silk, Chobani, etc. list sunflower lecithin ahead of gellan, which means it occurs at slightly higher weight than gellan but still < 0.5 % of the product; 0.04 ā 0.12 % is the range used in beverage R-&-D trials (Blue Diamond, PMC, Shine Additives).
Step 2 ā how that compares with doses that bother the gut
Emulsifier |
Human safety/tolerance benchmark |
Your maximum intake |
Margin |
Gellan gum (E 418) |
EFSA: āno ADI neededā; human volunteers tolerated 200 mg Ā· kgā»Ā¹ dayā»Ā¹ for 3 weeks without issues (European Food Safety Authority) |
0.85 g Ć· 70 kg ā 12 mg Ā· kgā»Ā¹ dayā»Ā¹
|
ā„ 17 Ć lower |
Xanthan / guar (if present) |
EFSA: transient GI bloating above 214 mg Ā· kgā»Ā¹ dayā»Ā¹ for 10 days (PMC) |
usually < 0.1 g total |
> 200 Ć lower |
Sunflower lecithin |
GRAS; typical background exposure ~55 mg Ā· kgā»Ā¹ dayā»Ā¹ in Western diets (Oxford Academic) |
2.4 g Ć· 70 kg ā 34 mg Ā· kgā»Ā¹ dayā»Ā¹
|
Within everyday range |
The human controlled-feeding trial that did show mucus-layer erosion used 15 g dayā»Ā¹ of carboxymethyl-cellulose (CMC) ā not gellan or lecithin ā for only 11 days (PMC). Your whole-day almond-milk quota delivers ~40 Ć less total emulsifier mass, and none of it is CMC or polysorbate-80, the two additives repeatedly implicated in the mouse colitis model (Nature, BioMed Central).
Step 3 ā mechanistic considerations
-
Gellan, xanthan & guar are high-Mw polysaccharides. They are not surfactants; they thicken the continuous (water) phase. They are minimally fermented and largely recovered intact in stool; no mucus-stripping activity has been demonstrated in vivo.
-
Sunflower lecithin is a phospholipid mix. Itās chemically similar to egg-yolk lecithin and is hydrolysed to choline + fatty acids by pancreatic phospholipase Aā long before it reaches the colon; again, no data link dietary lecithin to barrier erosion.
-
Synthetic surfactants (CMC, P80, sucrose fatty-acid esters, mono-diglycerides) can insert into micelles and bacterial envelopes; that is the class that triggers proximity of microbes to the epithelium in animal work. None of the mainstream unsweetened almond milks in the US market uses them.
Step 4 ā so is ½ā¾ gal a day āan issueā?
-
For the microbiome? Unlikely, provided your brand is free of CMC / polysorbate-80 / carrageenan. Your gellan + lecithin dose is well below levels considered safe in human and EFSA assessments, and mechanistic concern is low.
-
For nutrient balance? A bigger practical worry: drinking 2ā3 L of ultra-dilute almond milk crowds out protein (1 g Ā· cupā»Ā¹) and fiber; watch total calories and micronutrients.
-
Individual sensitivity exists. A minority of people in the CMC trial showed marked mucus thinning; if you have IBS/IBD and notice bloating or loose stools that track with almond-milk load, do an n = 1 trial off emulsifiers or switch to Elmhurst/MALK (no gums, no lecithin) or homemade.
How to minimise any residual risk
-
Scan the ingredient list ā avoid products that name carboxymethylcellulose, cellulose gum, polysorbate-80, mono- & diglycerides, or carrageenan.
-
Rotate brands ā spreads any additive exposure narrower and keeps the microbiome from adapting to one formulation.
-
DIY option ā 1 cup soaked almonds : 4 cups water + pinch salt; add ¼ tsp sunflower lecithin only if you hate shaking.
-
Diversify fiber elsewhere ā 25ā30 g/day of mixed soluble & insoluble fiber (oats, psyllium, legumes, veggies) buffers whatever trace gums you ingest and feeds butyrate producers.
Bottom line: At ½ā¾ gallon per day youāll take in roughly 1ā3 g of gellan + lecithin, a fraction of doses that raise red flags in either the toxicology or microbiome literature. Stay away from almond milks formulated with CMC or polysorbate-80 and the evidence says your gut barrier should be fine ā the bigger question is whether that much low-protein milk is crowding out more nutrient-dense foods.