Empanada: what is known is the oil and the dough

There is no US composition record for an empanada, so nothing on this page starts from a nutrition label. What has been measured is the oil it is fried in and the dough it is made of. In 19 healthy men, one high-fat meal made with repeatedly heated frying oil gave a smaller 4-hour rise in blood triacylglycerol than the same meal with unheated oil (p < 0.001), and flow-mediated dilation did not differ between the two. That was a single acute meal in men only, using a sunflower and palm olein blend heated for 10 days, so it says nothing about eating fried food for a year. A smaller four-hour rise in one blood fat is not a sign that reused oil is the safer choice. In 20 obese adults, muffins baked with sunflower oil that had been through 20 heating cycles raised three markers of endoplasmic reticulum stress inside blood cells (all p < 0.05), while olive oil and antioxidant-treated oils did not. Those are surrogate markers in mononuclear cells after one breakfast each. The wider signal about fried food is observational. Pooled across 17 studies, the highest fried-food intake went with 28 percent higher risk of major cardiovascular events than the lowest (RR 1.28, 95% CI 1.15 to 1.43), from self-reported intake with heterogeneity of 82 percent. The same body of work links the highest intake to 16 percent higher risk of overweight or obesity (RR 1.16, 1.07 to 1.25) and 20 percent higher hypertension risk (1.20, 1.05 to 1.38). It finds no clear link to type 2 diabetes (1.07, 0.90 to 1.27), and cross-sectional studies carry the obesity signal. Only one study here names the food. Among 1983 adults in two mid-sized Argentinian cities, a convenience and processed eating pattern built on empanadas, pizza and processed meat tracked with higher soluble E-selectin, a marker of endothelial dysfunction (P < 0.05). That is one time point, and it describes a whole pattern rather than the pastry.

How much is actually in it

• Acrylamide in 80 commercial biscuits sold in Spain averaged 343 micrograms per kg and ranged from under 20 to 2144 micrograms per kg, a spread of more than a hundredfold between products. — 80 commercial biscuits on the Spanish market, laboratory measurement by LC-ESI-MS/MS (n = 80 samples) Evidence D source
  Mean 343 microgram/kg, range below 20 to 2144 microgram/kg, rye, teff and oat highest, 70% of samples below the EU benchmark level

What cooking and storage change

• In 19 healthy men, a high-fat meal made with repeatedly heated frying oil lowered the 4-hour postprandial triacylglycerol rise compared with unheated oil, and flow-mediated dilation did not differ. — 19 healthy male participants, acute crossover, measurements over 4 h (n = 19) Evidence B source
  Postprandial TAG increase lower after heated oil (p < 0.001), FMD and arterial stiffness comparable between meals
• In 20 obese adults, muffins baked with sunflower oil that had gone through 20 heating cycles raised postprandial markers of endoplasmic reticulum stress in blood cells (all p < 0.05), while olive oil and antioxidant-treated oils did not. — 20 obese adults, four breakfasts in randomized crossover order (n = 20) Evidence B source
  Postprandial rise in sXBP1, BiP and CRT mRNA after sunflower-oil muffins (all p-values <0.05)
• A systematic review of food technology work reports that cutting asparagine and reducing sugars in the dough, or lowering baking temperature and time, limits how much acrylamide forms in carbohydrate-rich products. — food products and model systems, not people Evidence D source
  Direction only: lower asparagine, lower reducing sugars, lower temperature and shorter processing all reduce acrylamide formation, some additives increase it

What it does to your health

• Observational data pooled across 17 studies link the highest fried-food intake to 28 percent higher risk of major cardiovascular events than the lowest intake (RR 1.28, 95% CI 1.15 to 1.43). — general adult population, 17 studies for major cardiovascular events (n = 17 studies) Evidence C source
  RR 1.28 (95% CI 1.15 to 1.43) for major cardiovascular events, I2 82.0%, linear dose-response
• Pooled observational data link the highest fried-food intake to 16 percent higher risk of overweight or obesity (RR 1.16, 95% CI 1.07 to 1.25) and 20 percent higher risk of hypertension, with no clear link to type 2 diabetes. — adults, 11 studies for overweight or obesity, 12 for hypertension, 10 for type 2 diabetes (n = 11 studies for the obesity estimate) Evidence C source
  RR 1.16 (1.07-1.25) overweight or obesity, RR 1.20 (1.05-1.38) hypertension, RR 1.07 (0.90-1.27) type 2 diabetes
• Among 1983 adults in two Argentinian cities, a convenience and processed eating pattern built on empanadas, pizza and processed meat tracked with higher soluble E-selectin, a marker of endothelial dysfunction (P < 0.05). — 1,983 adults from two mid-sized Argentinian cities, biomarkers in a random subsample (n = 1,983) Evidence C source
  Higher adherence to the convenience and processed pattern directly associated with soluble E-selectin (P < 0.05)

Safety, limits and interactions

• A controlled 13-day feeding trial fed one meal containing an upper acrylamide exposure level of 14.1 to 15.9 micrograms per kg body weight, while four cups of coffee at 0.15 to 0.17 micrograms per kg raised urinary metabolites only within the background range. — healthy male volunteers on a duplicate diet for 13 days, acrylamide arm n = 6 (n = 6 in the acrylamide arm) Evidence B source
  Meal dose 14.1-15.9 microgram/kg bw as the upper exposure level tested, coffee dose 0.15-0.17 microgram/kg bw within background, endogenous baseline estimated at 0.3-0.4 microgram/kg bw per day
• EU law treats acrylamide as a carcinogenic substance that forms in cereal-based foods fried, roasted or baked above 120 degrees Celsius, and obliges food businesses to apply mitigation measures. — food businesses in the European Union, all cereal and potato based products Evidence E source
  Regulatory threshold of concern set at cooking temperatures above 120 degrees Celsius, benchmark levels per food category rather than legal maximum limits

What people believe that the data does not support

• Fourteen prospective cohorts found no significant link between dietary acrylamide intake and breast cancer (RR 0.95, 95% CI 0.90 to 1.01), endometrial cancer or ovarian cancer. — women in 14 prospective cohort studies (n = 14 cohort studies) Evidence C source
  Breast RR 0.95 (0.90 to 1.01), endometrial RR 1.03 (0.89 to 1.19), ovarian RR 1.02 (0.84 to 1.24)

Who this works differently for

• Across 53 870 mother and newborn pairs, the highest quartile of acrylamide haemoglobin adducts in pregnancy carried 1.20 times the odds of a small-for-gestational-age birth and 131 g lower birth weight. — 53,870 mother-newborn pairs from five cohort studies, exposure in pregnancy (n = 53,870 pairs) Evidence C source
  Odds ratio 1.20 (95% CI 1.08 to 1.33) for small for gestational age, birth weight -131 g (95% CI -204 to -58), head circumference -0.31 cm (95% CI -0.58 to -0.04)

The bottom line

Acrylamide is the compound people worry about in a browned pastry, so it is worth separating what has been measured from what is feared. It forms when cereal and potato foods are fried, roasted or baked above 120 degrees Celsius, which is the temperature the European Union writes into law when it obliges food businesses to apply mitigation measures. That is a regulator position rather than a study, and its benchmark levels are performance indicators rather than legal maximums. How much forms depends on the recipe. Across 80 commercial biscuits on the Spanish market the mean was 343 micrograms per kg, with a range from under 20 to 2144 and rye, teff and oat highest. Those were biscuits, not empanada dough, measured as sold rather than as baked in a home kitchen. Food technology work done in products and model systems, not in people, points one way: less asparagine and less reducing sugar in the dough, or a lower oven temperature and a shorter bake, mean less acrylamide forms. Then the cancer question. Fourteen prospective cohorts found no significant link between dietary acrylamide intake and breast cancer (RR 0.95, 95% CI 0.90 to 1.01), endometrial cancer (1.03, 0.89 to 1.19) or ovarian cancer (1.02, 0.84 to 1.24). Intake in those cohorts was estimated from food questionnaires, and a null result of that kind does not rule out a small effect. The reason the answer stays unsettled is that the dose people actually get has only been mapped in biomarker studies. A controlled 13-day feeding trial with six men in the acrylamide arm set the upper single-meal dose it tested at 14.1 to 15.9 micrograms per kg body weight. Four cups of coffee, at 0.15 to 0.17 micrograms per kg, moved urinary metabolites only within the background range, and those were biomarker endpoints that say nothing about disease risk. Who should be careful. The one population-specific finding here is pregnancy. Across 53 870 mother and newborn pairs, the highest quartile of acrylamide haemoglobin adducts carried 1.20 times the odds of a small-for-gestational-age birth (95% CI 1.08 to 1.33), 131 g lower birth weight and 0.31 cm smaller head circumference. That comes from five observational cohorts, with exposure taken from blood biomarkers and diet recall. We found no study of empanadas and children, allergy or medicines, and that gap is not a clearance. The larger hole is the food itself. There is no composition record under the name, no absorption study of this dough, and nobody has measured what an actual empanada, in an actual fryer, ends up containing.

Sources

1. empan-r5-08 · laboratory analysis of 80 retail biscuit samples
   The mean acrylamide content was 343 μg kg-1 (from <20 to 2144 μg kg-1).
   Food Funct, 2019 · checked 2026-09-24
2. empan-r5-01 · randomised crossover trial in humans
   The postprandial TAG increase was lower after heated versus unheated oil (p < 0.001).
   Lipids, 2026 · checked 2026-09-24
3. empan-r5-02 · randomised crossover trial in humans
   Our study showed that the intake of the muffins made with SFO induced the postprandial increase of the mRNA levels of the ER stress-sensor sXBP1, and the ER stress related chaperones BiP and CRT (all p-values <0.05).
   Mol Nutr Food Res, 2014 · checked 2026-09-24
4. empan-r5-09 · systematic review of food technology and in vitro studies
   Some interventions, such as reducing AA precursors in raw materials, (i.e., asparagine), reducing sugars, or decreasing temperature and processing time can be applied to limit AA formation in food products.
   Crit Rev Food Sci Nutr, 2024 · checked 2026-09-24
5. empan-r5-04 · meta-analysis of observational studies
   In comparing the highest with lowest fried-food intake, summary RRs (95% CIs) were 1.28 (1.15 to 1.43; n=17, I2=82.0%) for major cardiovascular events
   Heart, 2021 · checked 2026-09-24
6. empan-r5-05 · meta-analysis of observational studies
   In comparing the highest to lowest fried-food intake, the pooled RRs (95% CIs) were 1.16 (1.07-1.25; I2 = 71.0%, Pheterogeneity < 0.001) for overweight/obesity
   Crit Rev Food Sci Nutr, 2022 · checked 2026-09-24
7. empan-r5-10 · cross-sectional study in adults
   CDP consisted mainly of processed meat, snacks, pizza, and "empanadas", a stuffed bread served baked or fried...Contrariwise, higher scores in CDP were directly associated with sSELE concentrations (P < 0.05).
   Nutrition, 2019 · checked 2026-09-24
8. empan-r5-03 · controlled clinical trial in humans
   One study arm (group A, n = 6) ingested AA via coffee (0.15-0.17 µg/kg bw) on day 6 and in a meal containing an upper exposure level of AA (14.1-15.9 μg/kg bw) on day 10.
   Arch Toxicol, 2017 · checked 2026-09-24
9. empan-r5-11 · EU regulation summary, position of the European Commission
   Acrylamide is a carcinogenic substance that forms naturally when food, especially potato- or cereal-based products, coffee and coffee substitutes, is fried, roasted or baked at temperatures above 120°C.
   Reducing the presence of acrylamide in food, EUR-Lex summary of Commission Regulation (EU) 2017/2158 · checked 2026-09-24
10. empan-r5-07 · meta-analysis of observational cohort studies
    We found no significant association between dietary acrylamide intake and the risk of breast (relative risk (RR) 0·95; 95 % CI 0·90, 1·01), endometrial (RR 1·03; 95 % CI 0·89, 1·19) and ovarian cancers (RR 1·02; 95 % CI 0·84, 1·24).
    Br J Nutr, 2021 · checked 2026-09-24
11. empan-r5-06 · dose-response meta-analysis of observational cohort studies
    Higher risk of SGA and lower birth weight and head circumference were observed in the highest quartile of AA-Hb [odds ratio (OR): 1.20 (95% CI: 1.08; 1.33); mean difference (MD): -131 g (95% CI: -204; -58) and -0.31 cm (95% CI: -0.58; -0.04), respectively]
    Environ Res, 2022 · checked 2026-09-24