Movement & Power · Crete · GreecePublished 27 September 2026
What Did the 1948 Crete Food Survey Record?
A household survey recorded food in postwar Crete before the Seven Countries Study. Its dates, sample, and methods define what it can tell us.
One of the records used to describe older Cretan food has an unusually specific starting point: households visited by survey workers in 1948.
That date is useful.
It gives the evidence an address in time, before later summaries turn it into a general account of how people in Crete once ate.
The short answer
The survey recorded a diet dominated, in energy terms, by plant foods and oils and fats.
It also recorded the circumstances in which households lived. Leland G. Allbaugh's Crete: A Case Study of an Underdeveloped Area, published in 1953, arose from a Rockefeller Foundation investigation begun at the invitation of the Greek government.
Its purpose included understanding how living standards might be improved. Food was one part of that inquiry.
What was actually measured?
The report's methods distinguish a wider household sample from a smaller group supplying intensive dietary information.
The wider survey covered 740 households. For food inventories and dietary records, 128 households received daily visits over seven days from Greek Red Cross nurses.
Those numbers describe different layers of evidence. A statistic from the intensive food records should not be presented as though it came from every household in the wider survey.
The methods can be checked in the opening chapter of the digitized report preview, particularly printed pages 10–13.
What the food records show
A 2021 historical review in the British Journal of Nutrition, drawing on Allbaugh, summarizes the seven-day records as follows:
- Cereals, pulses, nuts, potatoes, vegetables, and fruit supplied approximately 61 percent of energy.
- Meat, eggs, fish, and dairy products supplied approximately 7 percent.
- Table fats and oils supplied approximately 29 percent.
These are shares of dietary energy, not percentages of food weight or the number of dishes served. The grouped categories also cannot tell us which individual food supplied most of any one household's meals.
The review is a later interpretation of the original records. It is identified here so that the route from the historical survey to these figures remains visible.
A week needs a season
Lin Foxhall and Hamish Forbes reproduce an Allbaugh dietary table in their 1982 study of grain as a staple food. The table identifies its underlying observations as one week in the fall season, covering 128 families.
That qualification belongs beside the numbers.
A measured week, an estimate of annual consumption, and a recollection of usual eating are different records. They may help check one another, but they should not be exchanged without explanation.
Before treating a figure as year-round intake, the relevant table and any seasonal adjustment need to be identified. The same applies to averages: they summarize households without reproducing the variation between them.
This was not the Seven Countries Study
The dates and research purposes differ.
The Seven Countries Study's institutional history places its Crete pilot work in 1957, followed by the formal international surveys beginning in 1958. Allbaugh's survey began in 1948 and investigated living conditions more broadly.
Treating the two as one investigation makes it harder to establish who was observed, what was recorded, and which conclusions belong to which evidence.
The companion note on Spain and the Seven Countries Study follows another instance where preliminary investigations and formal cohorts need to be distinguished.
What this record cannot settle
A household survey from 1948 cannot establish that the same diet had remained unchanged for centuries.
Nor do the food proportions alone demonstrate why people ate that way, whether they preferred it, or what health effect a particular ingredient produced. Those questions require additional evidence.
The record remains valuable within its limits. It gives us a dated observation that later descriptions can be checked against.
What remains uncertain
What did the full daily menus contain?
The accessible primary-source preview does not include the complete Food and Nutrition chapter or all its supporting tables. This note therefore avoids reconstructing a typical day's menu.
How did intake vary by season and household?
The one-week sample and later summaries cannot substitute for checking the report's full tables and adjustment procedures.
What did the surveyed people want to eat?
Food preference and satisfaction claims require their original passages and context. They are not inferred here from intake figures.
Source trail
Primary source
Leland G. Allbaugh, Crete: A Case Study of an Underdeveloped Area (Princeton University Press, 1953).
Foreword and opening methods chapter establish the survey's purpose and sample. The accessible preview does not contain the full chapter beginning on printed page 97.
Later scholarship
“Mediterranean Diet and cardiovascular health: an historical perspective,” British Journal of Nutrition (2021).
Source for the rounded energy shares summarized above, attributed by the review to Allbaugh's seven-day records.
Lin Foxhall and Hamish A. Forbes, “Sitometreia: The Role of Grain as a Staple Food in Classical Antiquity,” Chiron 12 (1982).
Table 5 reproduces Allbaugh's Table A49, identifying the one-week fall-season observations. This is a reproduction of historical data, not a new Cretan dietary survey.
Institutional chronology
Seven Countries Study, “History.”
Used solely to distinguish the later study's chronology from the 1948 household survey.
Research status
The survey's purpose and sampling framework are directly supported by the primary preview. Detailed dietary figures are attributed to later scholarship. Full menu reconstruction and claims about household preferences await inspection of the complete original passages.
About sourcing, uncertainty, and corrections: Research Standards