Are we adequately fuelled? Calories, body mass and the historical record

A calorie total needs context

Two people eating the same number of calories may be supporting very different bodies and daily demands. EPK uses energy in relation to body mass as one way to make that context visible. Population statistics offer a useful starting point: what do people report eating, how much do they weigh, and how is their intake divided between carbohydrate, fat and protein?

These questions deserve separate answers. Reported intake describes eating as captured by a survey. Measured expenditure describes energy use. Neither becomes the other simply because we divide by kilograms.

The modern US snapshot

For US adults aged 20 and older in August 2021–August 2023, USDA tables report average daily intakes of 2,330 kcal for men and 1,760 kcal for women. NCHS reports mean measured body masses of 90.3 kg and 77.9 kg respectively. Dividing those published means gives approximately 25.8 and 22.6 kcal/kg/day. [1,2]

These are illustrative ratios of group averages, not the average of each participant’s individual kcal/kg value. The dietary and body-measurement samples differ, and the body-mass table excludes pregnant women. A formal analysis should match records and eligibility before calculating individual ratios.

The same survey cycle’s published mean energy shares were 44.7% carbohydrate, 37.0% fat and 15.7% protein for men; for women, 45.9%, 37.8% and 15.3%. These are percentages of energy, not food weight. Alcohol and energy-calculation conventions mean the three figures need not total exactly 100%. [3]

What the figures suggest for EPK

A fat share in the upper 30s gives context to the contrast between common eating patterns and the lower-fat approach being investigated within EPK. It does not test whether changing that share improves outcomes. That requires intervention evidence.

The reported kcal/kg ratios also cannot establish a universal minimum. A value near 25 is not evidence that 25 is a biological threshold, nor that somebody below it is necessarily under-fuelled. Requirements vary, and the accuracy of the numerator matters.

The UK National Diet and Nutrition Survey provides a particularly relevant warning. Its 2019–2023 report describes approximately 30% overall under-reporting of energy intake, assessed against a doubly labelled water substudy. It is uncertain whether different macronutrients are misreported equally. This does not justify adding 30% to every food record, or applying a UK average correction to US participants. [4]

The famous US–Soviet comparison

A Reuters dispatch dated 8 January 1983, preserved in the CIA archive, attributed estimates of 3,280 calories per person per day in the Soviet Union and 3,520 in the United States to the CIA. The surviving item is a news clipping, rather than the underlying analytical report. [5]

The clipping also assigns 44% of Soviet calories to grains and potatoes, compared with 26% in the US. Its “fats and oils” category accounts for 17% and 18% respectively. Those categories describe foods: fat present in meat, eggs or dairy is not represented by the fats-and-oils category alone. It would therefore be incorrect to call either diet a 17–18%-fat diet. [5]

Before treating these historical estimates as intake measurements, we need the original methods: which losses were deducted, which population was covered and how consumption was estimated. We cannot infer that Americans previously ate 3,520 kcal but now eat much less by comparing this clipping with modern dietary recalls. The measurements are not demonstrated to be equivalent.

Individual surveys tell a different historical story

CDC’s analysis of US adults aged 20–74 reported age-adjusted intakes of 2,450 kcal for men and 1,542 kcal for women in 1971–1974. Carbohydrate, fat and protein supplied 42.4%, 36.9% and 16.5% for men, and 45.4%, 36.1% and 16.9% for women. These recall-based figures differ substantially from the headline national estimates. [6]

CDC explicitly notes recall bias and changes in interview methods when interpreting historical trends. This series is also age-adjusted and restricted to ages 20–74; the modern figures above cover adults aged 20 and older. It should not be presented as a directly matched before-and-after experiment. [6]

Dividing an all-population historical calorie estimate by an adult male body weight would introduce another mismatch. Historical kcal/kg comparisons are possible, but the population, period and measurement method need to agree.

From a useful question to a testable study

For EPK, the useful research question is whether intake relative to body mass and expenditure helps explain differences in daily function, hunger, recovery and dietary sustainability. A population average cannot answer that for an individual.

A next analysis could link dietary recalls with body measurements in the same survey participants, calculate survey-weighted individual kcal/kg values, and examine differences by age, sex and body composition. It should address reporting error and use independently measured expenditure where available. Associations from a cross-sectional survey would still not establish cause and effect.

The practical aim remains adequate nourishment that supports a life people can sustain. Historical comparisons can broaden the questions we ask; careful measurement is what allows us to answer them.

Related working papers

Read the relevant PDFs: EPK–001 to 007 are the revised version 1.2 papers; EPK–008 to 010 are version 1.0 supplements. Earlier PDFs and DOI records remain on the paper archive.