Salad dressing: the fat is the part that works

Salad dressing is usually where the fat on a salad comes from, and that fat is doing more than carrying flavour. In a randomised crossover trial in 12 women, absorption of the carotenoids and fat-soluble vitamins measured from a salad was highest with 32 g of soybean oil in the dressing (P < 0.002). The rise was linear across 0 to 32 g for alpha-carotene, lycopene, phylloquinone and retinyl palmitate. Responses varied widely between individuals in that trial and some subjects showed almost none, so the group average does not describe every reader. An earlier crossover in 7 healthy adults measured the other end of the same effect. After salads eaten with fat-free dressing, the appearance of alpha-carotene, beta-carotene and lycopene in chylomicrons was negligible, and the full-fat dressing beat the reduced-fat one (P < 0.02) as the reduced-fat beat the fat-free (P < 0.04). Seven people, twelve people and a single-meal blood measurement test absorption from that one meal: they say nothing about vitamin status over a year, and nothing about what 32 g of oil a day does to anything else. The fat-free bottle saves you the fat on the label and costs you part of the vegetables underneath it.

The numbers, per 100 g

What your body absorbs

• Salads eaten with fat-free dressing produced negligible carotenoid uptake into chylomicrons, while the same salad with 28 g of canola oil in the dressing produced the highest uptake. — 7 healthy adults, three salads in random order with 0, 6 or 28 g canola oil, blood sampled hourly for 12 hours (n = 7) Evidence B source
  Fat-free dressing: negligible appearance of alpha-carotene, beta-carotene and lycopene; reduced-fat higher than fat-free (P < 0.04); full-fat higher than reduced-fat (P < 0.02)
• Absorption of carotenoids and fat-soluble vitamins from a salad rose with the amount of soybean oil in the dressing across 0 to 32 g, and was highest at 32 g. — 12 women, five salads each with dressings containing 0, 2, 4, 8 or 32 g soybean oil, chylomicron AUC and Cmax measured (n = 12) Evidence B source
  Highest absorption of all carotenoids and fat-soluble vitamins at 32 g oil (P < 0.002); linear across 0 to 32 g for alpha-carotene, lycopene, phylloquinone and retinyl palmitate

What it does to your health

• Salad dressing carrying 3.6 g a day of plant sterol esters lowered LDL cholesterol by a further 9.7 percent and at the same time lowered total plasma carotenoids by 9.6 percent. — 26 men and 27 women on controlled isocaloric diets, two servings a day of ranch or Italian dressing with or without plant sterol esters, 3 weeks per diet (n = 53) Evidence B source
  LDL cholesterol -9.7 percent and triglycerides -7.3 percent with sterol esters; total plasma carotenoids -9.6 percent, values still within normal ranges
• Eating 20 g a day of soybean-oil mayonnaise instead of palm-olein mayonnaise for four weeks lowered total cholesterol by 0.25 mmol/L and LDL cholesterol by 0.17 mmol/L, and lowered HDL cholesterol by 0.12 mmol/L as well. — adults in a double-blind crossover trial, 20 g a day of each mayonnaise for 4 weeks with a 2-week washout Evidence B source
  Total cholesterol -0.25 mmol/L (P = 0.001), LDL-C -0.17 mmol/L (P = 0.016), HDL-C -0.12 mmol/L (P < 0.001), LDL-C:HDL-C ratio unchanged

Safety, limits and interactions

• Plant sterol esters delivered in reduced-fat spread and salad dressing at up to 9.0 g a day for eight weeks produced no product-related serious adverse events in 84 healthy adults. — 84 free-living healthy men and women, four arms at 0.0, 3.0, 6.0 or 9.0 g a day of phytosterol esters, eight weeks (n = 84) Evidence B source
  No difference in side effects or clinical laboratory values between arms; alpha- and trans-beta-carotene reduced in the 9.0 g arm versus control (P < 0.05), all carotenoid values within normal ranges
• The FDA-approved warfarin label says the amount of vitamin K in food may affect therapy and tells patients taking warfarin to keep the amount of vitamin K in their diet consistent. — patients taking warfarin sodium Evidence E source
  No effect size given; the label asks for a consistent amount of dietary vitamin K rather than avoidance

What people believe that the data does not support

• FDA lifts the total fat disqualifying level for salad dressings in its plant sterol and stanol ester heart-disease claim, so a dressing may carry that claim even though its fat content would block a health claim on any other food. — food labels in the United States Evidence E source
  No effect size; this is a regulatory carve-out written into the rule

Who this works differently for

• Preschoolers who taste bitterness strongly ate 80 percent more raw broccoli when salad dressing came with it as a dip or a sauce than when the broccoli was served plain. — 152 mostly Hispanic preschool children in Head Start classrooms, broccoli offered twice a week for 7 weeks in four dressing conditions (n = 152) Evidence C source
  +80 percent broccoli intake in bitter-sensitive children with dressing versus plain (P<0.001); no effect in bitter-insensitive children

The bottom line

Dressing is also the food that regulators picked as a carrier for cholesterol-lowering plant sterol esters. In a controlled feeding trial, 26 men and 27 women ate two servings a day of ranch or Italian dressing with or without 3.6 g of sterol esters, 3 weeks per diet. LDL cholesterol fell a further 9.7 percent and triglycerides 7.3 percent, and total plasma carotenoids fell 9.6 percent at the same time, with all carotenoid values still inside normal reference ranges, which makes that drop a measurable cost rather than a demonstrated harm. FDA lifted the total fat disqualifying level for salad dressings inside the same sterol claim rule, so a dressing may carry the heart-disease claim at a fat content that would block a health claim on any other food. That is a regulator deciding the evidence is sufficient, and it carries no effect size of its own. Changing which oil is in the jar does something small. Twenty grams a day of soybean-oil mayonnaise in place of palm-olein mayonnaise for four weeks lowered total cholesterol by 0.25 mmol/L (P = 0.001) and LDL cholesterol by 0.17 mmol/L (P = 0.016). It lowered HDL cholesterol by 0.12 mmol/L (P < 0.001) as well, and left the LDL to HDL ratio unchanged. The abstract of that crossover never reports how many people took part, and the comparison is one oil against another rather than against no mayonnaise. The plainest use of dressing has nothing to do with lipids. Among 152 mostly Hispanic preschool children in Head Start classrooms, the children who taste bitterness strongly ate 80 percent more raw broccoli when dressing came with it than when it was served plain (P<0.001). The children who do not taste bitterness that way ate no more. Classrooms rather than children were allocated to the conditions, so this sits below randomised evidence. Two cautions travel with the bottle. Plant sterol esters delivered in reduced-fat spread and salad dressing at up to 9.0 g a day for eight weeks produced no product-related serious adverse events in 84 free-living healthy adults. Alpha-carotene and trans-beta-carotene came out lower in the 9.0 g arm than in control (P < 0.05), with every value still in normal range. Eight weeks in healthy adults bounds short-term tolerability and leaves years of use, and children, untested. Separately, the FDA-approved warfarin label says the amount of vitamin K in food may affect therapy and asks patients to keep their vitamin K intake consistent rather than to avoid it. A label is a manufacturer text approved by a regulator and not a study, and a mention of vitamin K is no proof of a clinically important interaction with dressing. If you take warfarin and you are about to start pouring oil on a daily salad, raise it with whoever prescribes the drug.

Sources

1. dres-r4-01 · randomised crossover trial
   After ingestion of the salads with fat-free salad dressing, the appearance of alpha-carotene, beta-carotene, and lycopene in chylomicrons was negligible.
   Am J Clin Nutr, 2004 · checked 2026-09-24
2. dres-r4-02 · randomised crossover trial
   Absorption of all carotenoids and fat-soluble vitamins was highest with 32 g oil (P < 0.002).
   Am J Clin Nutr, 2017 · checked 2026-09-24
3. dres-r4-03 · randomised crossover feeding trial
   Total plasma carotenoids decreased 9.6% with PSE.
   Lipids, 2002 · checked 2026-09-24
4. dres-r4-08 · randomised crossover trial
   Daily use with 20 g of linoleic acid-rich SB-mayo elicited reductions in TC and LDL-C concentrations without significantly changing LDL-C:HDL-C ratio or small LDL particle distributions compared to the PO-mayo diet.
   Lipids Health Dis, 2016 · checked 2026-09-24
5. dres-r4-04 · randomised parallel trial
   The results of this study indicate that phytosterol esters are well tolerated and show no evidence of adverse effects at a daily intake of up to 9.0 g of phytosterols for eight weeks.
   J Am Coll Nutr, 2001 · checked 2026-09-24
6. dres-r4-05 · drug label
   The amount of vitamin K in food may affect therapy with warfarin sodium. Advise patients taking warfarin sodium to eat a normal, balanced diet maintaining a consistent amount of vitamin K.
   Warfarin sodium, FDA-approved label, drug interactions section, effective 2026-06-25 · checked 2026-09-24
7. dres-r4-06 · regulation
   All requirements set forth in § 101.14 shall be met, except § 101.14(a)(4) with respect to the disqualifying level for total fat per 50 grams (g) in dressings for salad and spreads and § 101.14(e)(6) with respect to dressings for salad.
   21 CFR 101.83, Health claims: plant sterol/stanol esters and risk of coronary heart disease (CHD) · checked 2026-09-24
8. dres-r4-07 · quasi-experimental trial
   Bitter-sensitive children, however, ate 80% more broccoli with dressing than when served plain (P<0.001)
   J Acad Nutr Diet, 2012 · checked 2026-09-24