Postprandial energy expenditure in whole-food and processed-food meals: implications for daily energy expenditure
Barr and Wright gave 17 healthy adults (12 women, 5 men) two isoenergetic cheese sandwiches in random order on separate days: a "whole-food" meal of multi-grain bread with sunflower seeds and whole kernels plus cheddar, and a "processed" meal of white bread plus a processed cheese product (600 or 800 kcal by label, chosen per person).
Oxygen uptake was measured hourly for 5–6 h. Energy spent digesting the whole-food meal was 137 ± 14.1 kcal, 19.9% of meal energy, against 73.1 ± 10.2 kcal, 10.7%, for the processed meal, nearly half as much; satiety did not differ.
The whole-food meal also had more protein (20% vs 15% of energy) and about three times the fibre, and the authors stress that the result applies to this meal only.
Protocol
18 recruited, one excluded for illness: n = 17 analysed (12 women, 5 men; 15 students aged 18–22 and 3 adults aged 47–56; BMI 22.0 self-reported).
Crossover: each ate both meals, randomised order, on consecutive days or up to a week apart, after a 12-h fast and no strenuous exercise that day; meals eaten within 20 min between 9:15 and 11:15.
Meals isoenergetic by label (600 or 800 kcal, same portion in both trials): whole-food 40% carbohydrate, 39% fat, 20% protein; processed 50% carbohydrate, 33% fat, 15% protein; whole-food meal ~3 times the fibre.
Measurement: two resting measures before the meal, then 2-min spirometry plus gas-bag oxygen sampling every hour for 5–6 h, seated; metabolic rate by a modified Weir equation with the respiratory quotient estimated from meal composition (0.831 vs 0.844), not measured. Thermic effect = area under the curve above resting rate.
Key Results
| Outcome | Whole-food meal | Processed meal | P (full text) |
|---|---|---|---|
| Energy spent on digestion (abstract, kcal) | 137 ± 14.1 kcal | 73.1 ± 10.2 kcal | — |
| Total thermic effect (Table 4, kJ, mean ± SE) | 576.3 (59) | 306.9 (42.6) | 0.0009 |
| Thermic effect as % of meal energy | 19.9 (2.5)% | 10.7 (1.7)% | 0.005 |
| Duration of the response | 5.8 h | 4.8 h | 0.001 |
| Satiety over 6 h | no difference | P = 0.78 | |
| Palatability (1–10) | 6.5 | 4.9 | 0.005 |
| Net energy gain from the meal (authors' calculation) | — | 9.7% higher for the processed meal | — |
Mechanisms
Measured: a larger and longer rise in oxygen uptake after the whole-food meal.
Authors' interpretation: whole foods are structurally more complex (intact grains, more fibre, more protein), so digesting them costs more; processed foods are easier to digest. The protein difference (5 percentage points) is, in their view, too small to explain the gap.
Inference, not tested: with oxygen sampled for 2 min an hour and the respiratory quotient assumed, the absolute values are uncertain; the 19.9% for the whole-food meal is about twice the ~10% of daily expenditure the authors themselves cite as typical for the thermic effect of food.
Contradicting evidence: in a 24-h respiratory chamber, ultra-processed vs unprocessed whole diets did not differ in sleeping or sedentary expenditure (REF-1130, whose authors contrast their result with this study).
Practical Application
A whole-food meal may cost more to digest, but this is one unconfirmed meal comparison: after one cheese-sandwich pair, 17 adults spent 137 vs 73.1 kcal digesting it (19.9% vs 10.7% of meal energy; B). The meals differed in protein and fibre as well as processing, the measurement was crude, and a chamber study of whole diets found no such gap (REF-1130). Choose whole food because it is easier to stop eating (REF-1130), not for a thermic bonus of this size.
For an athlete's daily count the effect is, at most, a few tens of kilocalories per meal and unproven at day level: the difference here, 64 kcal (arithmetic) for a 600–800 kcal meal, is inside ordinary logging error; do not adjust intake for it.
Do not generalise from this meal to "processed food" as a class: the authors restrict the finding to the meal tested.
Limitations
n = 17 young, mostly female adults, one test meal per condition, morning only, after a fast.
Protein (20% vs 15%), carbohydrate and fibre differed between meals, so processing is confounded with macronutrient content.
Energy contents taken from food labels; body mass and height self-reported.
Respiratory quotient assumed, oxygen sampled intermittently; one subject group only, no replication.
No day-level energy expenditure, no weight outcome.
How this entry was checked and changed
Why this level
a crossover with n = 17 would rate B+ by count; held at B because it is one meal pair, oxygen uptake was sampled for 2 min each hour with the respiratory quotient assumed from the menu rather than measured, the meals also differed in protein and fibre, and a 24-h chamber study of whole diets did not reproduce the gap (REF-1130); read from the full text
Full text (checked 2026-10-13)
n = 17 analysed (12 females, 5 males); 600 or 800 kcal portions; whole-food meal 40% carbohydrate, 39% fat, 20% protein; processed meal 50% carbohydrate, 33% fat, 15% protein (full text).
Total DIT 576.3 (59) vs 306.9 (42.6) kJ, P = 0.0009; DIT coefficient 19.9 (2.5)% vs 10.7 (1.7)%, P = 0.005; duration 5.8 vs 4.8 h, P = 0.001 (full text).
Satiety ratings did not differ over the monitoring period (P = 0.78); palatability 6.5 vs 4.9 (P = 0.005) (full text).
The DIT of the processed meal was 46.8% lower, giving a 9.7% increase in net energy gain for the processed meal (authors' calculation) (full text).
Respiratory quotient estimated from meal composition: 0.831 (whole food) and 0.844 (processed) (full text).