Almond calories: measured energy differed from the standard calculation
A controlled feeding study found lower metabolizable energy than standard factors predicted, without testing long-term weight loss.

Illustrative food image, not a photograph of the study intervention.
A question about measurement
This almond experiment asked how well a standard energy calculation matched energy measured in a human feeding study. It did not ask whether almonds prevented disease or produced lasting weight loss. Eighteen healthy adults consumed controlled diets in a crossover design, with almond amounts of 0, 42 or 84 g per day. Each treatment period lasted 18 days. The comparison was therefore unusually concrete: calculated energy values could be set beside values derived from the diets and the participants' excretion. The small, short experiment provides evidence about a food-energy estimate, not a complete account of the effects of eating nuts.
Novotny JA, Gebauer SK, Baer DJ. The American journal of clinical nutrition, 2012. PMID 22760558 · DOI
What the researchers collected
During the final nine days of each treatment period, volunteers collected all urine and feces. Samples of the diets, urine and feces were analyzed for energy and macronutrient content. Those measurements were used to determine the almonds' metabolizable energy within a mixed diet. This is a different question from simply listing the energy predicted by standard factors. The study's strength is the controlled feeding and collection procedure described in the abstract. Its limits are equally visible: only 18 healthy adults participated, and each dietary period lasted 18 days. The abstract does not establish a result for every almond preparation or every population.
Novotny JA, Gebauer SK, Baer DJ. The American journal of clinical nutrition, 2012. PMID 22760558 · DOI
Two numbers, one denominator
Measured almond energy was 4.6 ± 0.8 kcal per gram, equivalent to 129 kcal for a 28-g serving. The Atwater-factor estimate was 6.0–6.1 kcal per gram, or 168–170 kcal for the same serving size. The authors describe the standard factors as overestimating the measured value by 32%. That percentage uses the measured value as the reference; it should not be casually rewritten as 32% fewer calories than the prediction. The serving size also matters. These are comparisons for equal amounts of almonds, not comparisons between a small almond portion and an unrestricted serving of another food.
Novotny JA, Gebauer SK, Baer DJ. The American journal of clinical nutrition, 2012. PMID 22760558 · DOI
Not a weight-loss prescription
The authors conclude that standard factors can be inaccurate for certain applications and describe a method for measuring an individual food's energy within a mixed diet. Their abstract does not report a long-term weight-loss endpoint. It would therefore overreach to promise that adding almonds causes weight loss or to deduct a fixed amount of energy from every almond-containing recipe. The accompanying almond porridge is a culinary pairing, not the controlled diet used in the trial. The evidence supports a narrower and useful conclusion: in this experiment, measured metabolizable energy from almonds was lower than the value predicted by the standard calculation.
Novotny JA, Gebauer SK, Baer DJ. The American journal of clinical nutrition, 2012. PMID 22760558 · DOI