Flan: a caramel custard that reaches the database as a powder
Flan is a baked caramel custard, and in the United States reference tables it arrives as a packet: all three records are the dry mix, on its own or made up with 2 percent or whole milk. The research sits at a similar distance. The nearest measurement is about structure. Twelve healthy adults in a randomised open label crossover ate four foods matched on composition. The triglyceride response after the custard ran up to 122.7 percent higher than after the pudding and the biscuit (p < 0.0001). That was a single postprandial test on model foods rather than on a baked flan. A dessert that is soft is also a dessert that keeps going down. Twenty healthy adults with a mean age of 23 ate 66 plus or minus 21 g less chocolate custard in one sitting when it was made to stay longer in the mouth. In the companion trial, 40 adults ate 42 plus or minus 15 g less when they slowed the pace down, and that happened only at the longer mouth exposure (p = 0.04).
Every statement here is tied to a source. The badge says what kind of evidence stands behind it: A is a meta-analysis or systematic review, E is the position of an expert body without its own analysis. How we verify →
The numbers, per 100 g
| USDA entry | kcal | Protein | Carbs | Fiber |
|---|---|---|---|---|
| Flan, caramel custard, dry mix | 348 | 0 g | 91.6 g | 0 g |
| Flan, caramel custard, dry mix, prepared with 2% milk | 103 | 2.99 g | 18.8 g | 0 g |
| Flan, caramel custard, dry mix, prepared with whole milk | 113 | 2.95 g | 18.7 g | 0 g |
These are the USDA entries that are flan, not foods made from it. Follow a name for the full profile and per-portion figures.
What a real portion looks like
- Healthy young adults ate 66 g less chocolate custard in one sitting when the same dessert was made to stay longer in the mouth. — 20 healthy adults, mean age 23 y, BMI 23, eating chocolate custard until they felt full (n = 20) Evidence B source66 g less intake (SEM 21) at high oro-sensory exposure compared with low
- Slowing the pace of eating cut custard intake by 42 g, and only when each bite was kept in the mouth longer. — Participants of study 1, eating chocolate custard until fullness under four randomised conditions (n = 40) Evidence B source42 g less intake (SEM 15) in the slow eating rate, high oro-sensory exposure condition compared with fast eating rate at the same exposure, P = 0.04
What your body absorbs
- Fat left a custard into the blood far faster than from a pudding or a biscuit of the same composition, with a triglyceride response up to 122.7% higher. — Twelve healthy subjects in a randomized, open label, crossover postprandial trial (n = 12) Evidence B sourceTriglyceride chylomicron response up to 122.7% higher after custard than after pudding and biscuit, p < 0.0001
What it does to your health
- At 24 hours the gut bacteria of the ten adults who ate one 192 g serving of low glycaemic index custard cake differed from those of the high glycaemic index arm, and the trial says such effects may be temporary. — 30 healthy individuals aged 18 to 45 years, 10 per dessert group, one serving alongside a standard breakfast, stools sampled at baseline and 24 h (n = 30) Evidence B sourceDirection only: Collinsella and Bifidobacterium lower in the high glycaemic index group than in the low and medium groups, no effect size reported
Safety, limits and interactions
- Bacillus cereus turned up in 45.9% of 381 dried milk dessert samples, flan among them, at 3.0 to 10000 spores per gram. — 381 samples of dried milk products including flan, used by the Chilean School Feeding Program (n = 381 samples) Evidence C source175 of 381 samples positive (45.9%), levels from 3.0 to 10(4) spores per g, and 28 of 94 tested isolates (29.8%) produced diarrhoeal enterotoxin
- The FDA tells cooks to bring egg dishes such as casseroles to 160 F, because a clean uncracked shell egg can still carry Salmonella. — General public cooking dishes that contain eggs Evidence E sourceTarget internal temperature 160 F for egg-containing dishes
What people believe that the data does not support
- The cancer warning that follows the word caramel on some drinks is about 4-methylimidazole formed while manufacturing caramel colour, a beverage colourant. — 12 beverages bought in California and the New York metropolitan area, with US intake taken from NHANES (n = 12 beverages, 110 samples) Evidence C sourceMeasured 4-MEI 915.8 to 963.3 µg/L in the highest beverage and 9.5 to 11.7 µg/L in the lowest, lifetime excess cancer risk 1.93x10-4 to 1.92x10-6
- Among 105260 French adults, the heaviest consumers of plain caramel colour E150a had 15% higher overall cancer incidence than non or low consumers. — 105,260 adults without prevalent cancer, 78.3% women, mean age 42.0 y, followed for more than 7 years in the NutriNet-Sante cohort (n = 105260) Evidence C sourceHR 1.15 (95% CI 1.07 to 1.25) for overall cancer, absolute risk at age 60 of 14.0% against 12.1%
The bottom line
The one study on this page that names flan is about the powder. Of 381 samples of dried milk products used by the Chilean school feeding programme, flan among them, 175 (45.9 percent) carried Bacillus cereus at 3.0 to 10000 spores per gram, and 28 of the 94 isolates tested (29.8 percent) produced diarrhoeal enterotoxin. Those are food samples from one national programme, and nobody's illness was measured. The other safety line comes from the eggs. The United States Food and Drug Administration says a clean uncracked shell egg can still carry Salmonella and that dishes containing eggs should reach 160 F, on a page that names casseroles, quiches and souffles without naming flan. Then there is the caramel, which is where the word does most of its damage. The cancer warning that follows caramel on some drink labels is about 4-methylimidazole, a compound formed while caramel colour is manufactured as a beverage colourant. Measured across 12 beverages and 110 samples, it ran from about 9.5 to about 963.3 micrograms per litre. In a French cohort of 105 260 adults followed more than 7 years, the heaviest consumers of plain caramel colour E150a had 15 percent higher overall cancer incidence than non or low consumers. The hazard ratio was 1.15 (95% CI 1.07 to 1.25), an absolute risk at age 60 of 14.0 percent against 12.1. That cohort was 78 percent women and volunteer-recruited, and its authors call the additive analysis exploratory. None of this work measured sugar melted in a pan. The remaining study is a third dessert again. Thirty healthy adults aged 18 to 45, ten per arm, ate one Thai dessert with a standard breakfast. In the 192 g custard cake arm the gut bacteria differed from the high glycaemic arm at 24 hours. That is a direction-only result with no effect size, and the trial itself says such shifts may be temporary. The cake in question is a coconut and egg custard from Phetchaburi. What is absent is worth naming plainly. Nothing on this page speaks about pregnancy, about children, about egg or milk allergy, or about any medicine, so it cannot tell you who should avoid a flan.
Sources
- flan-daily-r3-01 · randomised controlled trial
In study 2, participants ate 66 ± 21 g less in the high- than in the low-OSE condition and there were no intake differences between slow and fast ER (P = 0.35).
Am J Clin Nutr, 2020 · checked 2026-09-24 - flan-daily-r3-02 · randomised controlled trial
Participants ate significantly less (42 ± 15 g) in the slow-ER, high-OSE than in the fast-ER, high-OSE condition (P = 0.04).
Am J Clin Nutr, 2020 · checked 2026-09-24 - flan-daily-r3-03 · randomised controlled trial
Custard triglyceride response is higher than pudding and biscuit responses (up to +122.7%, p < 0.0001).
Mol Nutr Food Res, 2020 · checked 2026-09-24 - flan-daily-r3-04 · randomised controlled trial
Participants were randomly assigned to one of three groups: Phetchaburi's Custard Cake (192 g, low-GI group, n = 10), Saraburi's Curry Puff (98 g, medium-GI group, n = 10), and Lampang's Crispy Rice Cracker (68 g, high-GI group, n = 10), each consumed alongside their standard breakfast...The findings suggest that Thai desserts with varying GI levels can influence specific gut bacteria, though these effects may be temporary.
Nutrients, 2024 · checked 2026-09-24 - flan-daily-r3-05 · prevalence survey
B. cereus was found in 175 of 381 of the samples analysed (45.9%), reaching levels from 3.0 to 10(4) spores g(-1).
Food Microbiol, 2007 · checked 2026-09-24 - flan-daily-r3-06 · agency guidance
Fresh eggs, even those with clean, uncracked shells, may contain bacteria called Salmonella that can cause foodborne illness...Casseroles and other dishes containing eggs should be cooked to 160° F.
U.S. Food and Drug Administration, What You Need to Know About Egg Safety · checked 2026-09-24 - flan-daily-r3-07 · exposure and risk assessment
Caramel color is added to many widely-consumed beverages as a colorant. Consumers of these beverages can be exposed to 4-methylimidazole (4-MEI), a potential carcinogen formed during its manufacture.
PLoS One, 2015 · checked 2026-09-24 - flan-daily-r3-08 · prospective cohort
After False Discovery Rate correction, only plain caramel (European code: E150a) was associated with overall cancer [1.15 (1.07-1.25); 14.0%, 12.1%] and beta-carotene (E160a) with overall [1.16 (1.07-1.25); 13.7%, 11.9%] and breast cancer [1.41 (1.23-1.62); 6.2%, 4.4%].
Eur J Epidemiol, 2026 · checked 2026-09-24
Review. Reviewed on 2026-09-24 for evidence level, dose and population context, and claims a reader could misread. Bioma Learn has no human medical reviewer at this time, and we say so rather than invent one. Methodology. This is information, not medical advice.