Sandwich: one word covering a threefold range

Nobody can tell you what is in a sandwich, because a sandwich is an assembly rather than a food. USDA has no composition record whose subject is a sandwich at all. The UK food table has one row: a white bread ham salad sandwich at 163 kcal per 100 g, with 8.2 g of protein, 25.0 g of carbohydrate and 639 mg of chloride. That is one filling on one bread, and there is no second dataset to check it against. The word is just as loose about size. The US food survey weighs a regular sandwich at 115 g, a large one or a six inch submarine at 175 g and a slider at 60 g, all of them reporting conventions for dietary recalls. American labelling law files all types of sandwiches under mixed dishes, with a reference amount of 140 g per eating occasion. What research can measure is how the same lunch gets eaten. Serving 69 healthy adults ad libitum sandwich meals, the slow and low-density version stopped at 570 kcal (95% CI 442 to 698) while the fast and energy-dense version reached 1143 kcal (95% CI 1015 to 1271). Those sandwiches were built to hit energy densities of 1.9 and 3.8 kcal/g and target eating rates, in a laboratory lunch with young adults of normal weight.

How much is actually in it

• The UK food table gives a white bread ham salad sandwich 163 kcal per 100 g with 8.2 g of protein, 4.1 g of fat, 25.0 g of carbohydrate and 639 mg of chloride, and USDA holds no sandwich composition record to check it against. — CoFID 2021 record Sandwich, white bread, ham salad, code 11-965, group AB; USDA FoodData Central has no composition entry whose subject is a sandwich (n = 1 record) Evidence E source
  Per 100 g: energy 163 kcal and 690 kJ, protein 8.2 g, fat 4.1 g of which 0.94 g saturated, carbohydrate 25.0 g with 22.5 g starch and 2.5 g total sugars, AOAC fibre 1.6 g, water 58.5 g, chloride 639 mg, calcium 98 mg, iron 1.04 mg, selenium 6 mcg

What a real portion looks like

• The US food survey weighs a regular sandwich at 115 g, a large one or a 6 inch submarine at 175 g and a slider at 60 g, so one word covers a threefold range of portion. — USDA FNDDS survey food Sandwich, NFS, WWEIA category Deli and cured meat sandwiches, FNDDS code 27500050, five household measures (n = 5 household measures) Evidence E source
  1 regular 115 g, 1 large 175 g, 1 regular or 6 inch submarine 175 g, 1 miniature or slider 60 g, quantity not specified 115 g
• US labelling law files all types of sandwiches under mixed dishes not measurable with a cup, with a reference amount of 140 g per eating occasion and 55 g added for a product carrying gravy or sauce. — 21 CFR 101.12 Table 2, reference amounts customarily consumed per eating occasion, general food supply, mixed dishes category (n = 1 regulatory category) Evidence E source
  Reference amount 140 g for all types of sandwiches, 195 g with gravy or sauce topping; label statement expressed as pieces or fractional slices with the gram weight

What your body absorbs

• In 11 fasted men, true zinc absorption from a chicken sandwich did not differ significantly between white and wholemeal bread, 33.6 percent against 25.4 percent. — eleven fasted adult men, a chicken meat sandwich on white or wholemeal bread labelled with 2 mg of zinc-67 on two separate occasions, double-label stable isotope technique with an intravenous zinc-70 dose (n = 11) Evidence B source
  Mean true zinc absorption 33.6% from the white bread sandwich and 25.4% from the wholemeal one, no significant difference; true absorption ran about 7% higher than apparent absorption

What it does to your health

• Serving 69 adults ad libitum sandwiches, the slow and low-density version stopped at 570 kcal while the fast and high-density version reached 1143 kcal, a gap of 573 kcal at one lunch. — 69 healthy adults, mean age 25 and BMI 23, five ad libitum lunchtime sandwich meals in a 2 by 2 randomised crossover with a medium control, sandwiches carrying 67 to 98 energy percent ultra-processed ingredients (n = 69) Evidence B source
  Slow eating rate with low energy density 570 kcal (95% CI 442 to 698) against fast rate with high density 1143 kcal (95% CI 1015 to 1271); versus the typical-lunch control, -59% (394 kcal) and +19% (179 kcal)
• Two high-fat sandwich meals carrying 40 g of milk fat each produced the same postprandial triglyceride rise in 12 healthy adults whether the milk fat globule membrane was intact, disrupted or absent (P = 0.9). — 12 healthy adults, randomised double-blind crossover, three visits inside whole-room indirect calorimeters at least a week apart, two high-fat sandwich meals per visit with 60 g of dairy spread each including 40 g of milk fat (n = 12) Evidence B source
  Triglyceride incremental AUC 179.5 (95% CI 76.9 to 282.2), 155.0 (83.5 to 226.5) and 133.4 (75.2 to 191.6) mmol/L/min, P = 0.9; insulinotropic polypeptide rose more with intact membrane (P = 0.02)

Safety, limits and interactions

• Across 27 806 food samples in 59 studies, 11 percent held more than 20 mg/kg of gluten and 0.51 percent passed 200 mg/kg. An ordinary 100 to 300 g portion of a food above 200 mg/kg could break the 10 mg a day safety threshold. — 59 studies from 28 countries, 1993 to 2024, 27 806 real-life food samples analysed predominantly by sandwich ELISA R5; oats showed the highest average contamination and restaurant meals the greatest inconsistency (n = 27 806 samples in 59 studies) Evidence C source
  3100 samples (11%) above 20 mg/kg, 359 (1.3%) above 80 mg/kg, 142 (0.51%) above 200 mg/kg; a conservative pooling of 15 467 samples puts 20% above 20 mg/kg
• In a scoping review of 40 studies of ready-to-eat street food in Africa, sandwiches were among the most frequently reported carriers of E. coli, Salmonella and Staphylococcus, and 10 of the papers reported 54 antimicrobial resistance genes. — 40 published prevalence studies of ready-to-eat street foods in Africa, population to case ratio 11 653 to 5 338 (45.80%) (n = 40 studies) Evidence C source
  Sandwiches named among the most frequently reported contaminated ready-to-eat foods; 54 resistance genes reported by 10 authors; 48 antibiotics in use against the infections

What people believe that the data does not support

• In a real restaurant, 454 UK diners handed menus with red triangle salt warnings were far more likely to think about the salt in their meal than diners with the standard menu, odds ratio 19.50 (95% CI 8.24 to 46.16). — study 2: 454 UK adults who eat out at least monthly, block randomised to a menu carrying red triangle salt warnings next to high-salt items (n = 240) or the restaurant's standard menu (n = 225), buying and eating a real lunch; study 1: 2391 UK adults online, recruited among people who buy supermarket sandwiches and savoury snacks (n = 2391 online and 454 in the restaurant) Evidence B source
  Perceived message effectiveness difference 1.00 (95% CI 0.79 to 1.18) for the labelled menu; odds ratio 19.50 (95% CI 8.24 to 46.16) for thinking about salt content when ordering; online arm 1.23 (95% CI 1.12 to 1.34) for packaged food scenarios
• Among 6210 US adults ordering lunch from two fast food menus, traffic light ecolabels nudged the nutrition score up in the burger restaurant and did nothing at all in the sandwich restaurant, while cutting that sandwich meal's greenhouse gas emissions by 29.8 percent. — 6210 US adults aged 18 and over from a nationally representative panel, randomised to one of five menu label conditions, ordering lunch from a burger-focused and a sandwich-focused fast food menu online in April 2025 (n = 6210) Evidence B source
  Nutrient Profile Index 49.93 against 49.25 in the burger restaurant (P < .001) and no ecolabel effect in the sandwich restaurant; greenhouse gas emissions 15.2% lower in the burger restaurant and 29.8% lower in the sandwich restaurant under traffic light labels

The bottom line

Two questions people ask about sandwiches are settled less firmly than they sound. In 11 fasted men in 1992, true zinc absorption from a chicken sandwich was 33.6 percent on white bread and 25.4 percent on wholemeal. That gap of 8 percentage points did not reach significance, and eleven men is too few to decide whether the breads differ. Two high-fat sandwich meals carrying 40 g of milk fat each raised triglycerides the same way in 12 healthy adults, whether the milk fat globule membrane in the spread was intact, disrupted or absent (P = 0.9). There the sandwich was a delivery vehicle for a purpose-made spread, and the load was 80 g of milk fat across two sandwiches at one sitting. Menu labels do less than they promise. When 454 UK diners in a real restaurant were handed menus carrying red triangle salt warnings, they were far more likely to think about the salt in their meal, odds ratio 19.50 (95% CI 8.24 to 46.16). The outcome measured there was attention rather than salt eaten, which the authors say themselves. Among 6210 US adults ordering lunch from two fast food menus online, traffic light ecolabels lifted the nutrition score by 0.68 of a point out of 100 in the burger restaurant. In the sandwich restaurant they had no effect at all. The same labels cut that sandwich meal's greenhouse gas emissions by 29.8 percent. These were orders placed on a survey platform rather than food eaten. The cautions follow from the assembly. A sandwich is several ready-to-eat parts put together with no final cooking step. That is why it appears in a scoping review of 40 studies of street food in Africa, among the foods most often reported carrying E. coli, Salmonella and Staphylococcus. Ten of those papers reported 54 antimicrobial resistance genes between them. Prevalence in street food in resource-limited settings does not transfer to a chilled supermarket pack. Nothing here measures a chilled supermarket pack either, so that sandwich is neither counted nor cleared. On gluten, a scoping review of 27 806 real-life food samples found 11 percent above 20 mg/kg and 0.51 percent above 200 mg/kg. An ordinary 100 to 300 g portion of a food above 200 mg/kg could break the 10 mg a day safety threshold, on the authors' own crude pooling of foods in general rather than of sandwiches, and the review does not say that the samples below 200 mg/kg are clear. Restaurant meals were the most inconsistent source in that review, which is the part that matters when you order a gluten-free sandwich out. We found no study of who a sandwich works differently for, and nothing in this material speaks to allergy, children, pregnancy or medicines.

Sources

1. sndw-r3-08 · food composition dataset
   Sandwich, white bread, ham salad, code 11-965, Energy 163 kcal, Protein 8.2 g, Fat 4.1 g, Carbohydrate 25.0 g, Chloride 639 mg per 100 g
   CoFID 2021, Public Health England, OGL v3.0, code 11-965 · checked 2026-09-24
2. sndw-r3-09 · food survey portion dataset
   Sandwich, NFS, 1 regular 115 g, 1 large 175 g, 1 miniature/slider 60 g, FNDDS code 27500050, fdc_id 2706880
   USDA FNDDS, Sandwich NFS, fdc_id 2706880 · checked 2026-09-24
3. sndw-r3-10 · regulation, 21 CFR
   Not measurable with cup, e.g., burritos, enchiladas, pizza, pizza rolls, quiche, all types of sandwiches...140 g, add 55 g for products with gravy or sauce topping
   21 CFR 101.12, Reference amounts customarily consumed per eating occasion, Table 2 · checked 2026-09-24
4. sndw-r3-01 · randomised crossover isotope study in men
   There was no significant difference in mean true Zn absorption from the white or wholemeal bread sandwich, 33.6 and 25.4% respectively. It was concluded that the substitution of wholemeal for white bread does not reduce Zn absorption from meat-based sandwiches.
   Br J Nutr, 1992 · checked 2026-09-24
5. sndw-r3-02 · randomised crossover feeding trial
   Energy intake of the slow-ER/low-ED meal (mean 570 kcal; 95% confidence interval: 442, 698 kcal) was 573 kcal (50%) lower compared to the fast-ER/high-ED meal [1143 kcal; (1015, 1271 kcal)].
   J Nutr, 2025 · checked 2026-09-24
6. sndw-r3-07 · randomised crossover feeding trial
   The presence or structural integrity of MFGM in a novel high-fat spread incorporated into 2 standardized sandwich meals does not acutely influence postprandial lipidemia in healthy individuals.
   J Nutr, 2026 · checked 2026-09-24
7. sndw-r3-03 · scoping review of observational contamination data
   Of these, 3100 (11 %) had a gluten content exceeding 20 mg/kg, 359 samples (1.3 %) contained >80 mg/kg, and 142 samples (0.51 %) > 200 mg/kg...Regular portions (100-300 g) of foods >200 mg/kg could exceed the 10 mg/day safety threshold
   Clin Nutr ESPEN, 2026 · checked 2026-09-24
8. sndw-r3-04 · scoping review of observational prevalence studies
   The most frequently reported RTE foods were meat or beef/beef-soup, chicken or poultry products, salads, vegetable salads, and sandwiches, which harboured pathogens such as E. coli, Salmonella, and Staphylococcus.
   Front Microbiol, 2025 · checked 2026-09-24
9. sndw-r3-05 · randomised controlled trials in adults
   Participants assigned to the labelled menu condition were significantly more likely to think about the salt content of the meals when ordering than were participants assigned to the standard menu (odds ratio 19·50, 95% CI 8·24-46·16; p<0·0001).
   Lancet Public Health, 2025 · checked 2026-09-24
10. sndw-r3-06 · randomised clinical trial
    Only the TLL was effective at promoting more healthful food choices in the burger restaurant compared to the control label (mean [standard error] NPI score: 49.93 [0.18] points vs 49.25 [0.18] points; P < .001). There were no ecolabel effects on NPI score in the sandwich restaurant.
    JAMA Health Forum, 2026 · checked 2026-09-24