What starvation research reveals about restrictive dieting

Beyond the number on the scales

A restrictive diet can change more than body mass. Keys and colleagues examined energy expenditure, physical capacity, appetite, mood and recovery. For EPK, the central question is whether an eating pattern supports a functioning life and can be sustained without persistent physiological and psychological strain.

The passages below come from the two volumes of The Biology of Human Starvation (1950). The book combines Minnesota findings with earlier experiments, famine reports and clinical observations. Each passage is explained in that context. The opening two passages are from Russell Wilder’s foreword.

Original wording is retained with typographical corrections; […] indicates omitted text. Two entries are identified as table summaries. Historical findings inform the research questions here; they do not establish that every calorie deficit is harmful or validate a universal intake threshold.

p. xviii | Metabolism and measurement

Russell M. Wilder, foreword in Keys et al., Volume 1, p. xviii.

It is interesting that the obese man or woman can be placed on a diet as low as 600, or even 400, calories without loss of body nitrogen or other ill effects and without depression of the basal metabolic rate. In these respects obese persons and persons of normal weight differ strikingly in their response to a low intake of calories.

Relevance to EPK and restrictive dieting: Wilder’s foreword presents a historical contrast between obesity and normal-weight starvation. For EPK, it argues for considering body stores and clinical context rather than applying one response to everyone. The historical claim is not a modern demonstration that 400–600 kcal/day diets preserve all aspects of health.

p. xix | Metabolism and measurement

Russell M. Wilder, foreword in Keys et al., Volume 1, p. xix.

The basal metabolic rate in semi-starvation may fall as much as 50 per cent below normal standards based on energy exchange per square meter of surface area.

Relevance to EPK and restrictive dieting: The passage concerns a fall in basal expenditure normalised to body surface area. It supports investigating physiological adaptation during severe restriction, but it cannot be converted directly into an intake-per-kilogram minimum or treated as a 50% fall in every person's total daily needs.

p. xxiii | Physical function and clinical context

Authors’ foreword, Keys and colleagues, Volume 1, p. xxiii.

And, whether the origin is primary or secondary the presence of undernutrition makes a special kind of person, different morphologically, chemically, physiologically, and psychologically from his well-fed counterpart.

Relevance to EPK and restrictive dieting: The whole-person description fits EPK's proposal to assess function and behaviour alongside weight. The relevant criticism is that weight loss alone may conceal deterioration; the passage does not specify a validated readiness score or show that every deliberate deficit has these effects.

p. 15 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 15.

In the individual the severity of clinical edema is not by itself a reliable guide to his nutritional state. Where the complications of malaria, dysentery, anemia, and vitamin deficiencies exist, the edema appears more readily and in more extreme form.

Relevance to EPK and restrictive dieting: Fluid retention can obscure the relation between scale weight and tissue depletion. EPK assessments should consider clinical context rather than interpret a plateau as evidence that restriction is insufficient. The passage itself identifies infections and deficiencies as additional causes.

p. 23 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 23.

The substantial improvement of the diet in 1943 was paralleled by an epidemic incidence of hypertension.

Relevance to EPK and restrictive dieting: The reported parallel between improved food supply and hypertension complicates any claim that refeeding is uniformly beneficial. This is a historical association requiring context, not evidence that eating more directly causes hypertension or that prolonged restriction is protective.

p. 40 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 40.

Table 2 summary: six weeks after the end of diet restriction, reported recovery weights in Squad A exceeded control weights by 0.7 to 7.2 kg. These are Carnegie Nutrition Laboratory data compiled from Benedict and colleagues (1919), not Minnesota participants.

Relevance to EPK and restrictive dieting: The recovery overshoot is evidence from an earlier restriction experiment discussed by Keys, rather than an additional Minnesota result. It motivates tracking recovery beyond the lowest body mass. A six-week weight comparison cannot show permanent fat gain or predict the result of every diet.

p. 45 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 45.

The increased sensitivity to cold was evidenced in several ways in addition to the subjective reports: nearly all the men wore heavier underclothing and overclothing.

Relevance to EPK and restrictive dieting: Cold intolerance is a plausible functional sign worth tracking during under-fuelling. It challenges evaluating a restrictive diet solely by weight change, but coldness is nonspecific and cannot diagnose an energy deficit or establish a calorie prescription.

p. 55 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 55.

the red blood cell count dropped from a pre-starvation level of 5.52 to 4.45 million per cu. mm. during semi-starvation, and the hemoglobin value from 87 to 81 per cent.

Relevance to EPK and restrictive dieting: A fall in red-cell measures suggests that physical capacity may deteriorate alongside weight during severe restriction. This supports broader outcome monitoring in EPK; it does not isolate calorie shortage from nutrient status, fluid changes or other explanations for anaemia.

p. 62 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 62.

The men were commonly found to be somewhat irritable during the semi-starvation period. All the subjects reported a distinctly decreased sex interest and a reduced frequency of erections while on the restricted diet.

Relevance to EPK and restrictive dieting: Irritability and diminished sexual interest connect energy restriction with wellbeing and reproductive function. They support investigating whether a diet is becoming physiologically costly, rather than interpreting distress as weak commitment. Neither symptom is specific to low intake.

p. 78 | Appetite, tissue recovery and refeeding

Keys and colleagues, Volume 1, p. 78.

Extremely high food intakes […] as high as 7000-10,000 Cal. per day […] were achieved by most of the subjects during the first 2 weeks of ad lib eating, but the intakes tended to level off at more reasonable values of about 3200 to 4500 Cal. per day.

Relevance to EPK and restrictive dieting: Very high intake during early recovery is consistent with powerful post-deprivation eating pressure. EPK can investigate this as a recovery response rather than assume lack of restraint. These historical amounts are not targets for contemporary dieters or evidence that all increased eating is beneficial.

p. 87 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 87.

According to this report the average actual caloric intake in 44 factories in the Tokyo area steadily declined from a level of 3012 Cal. daily in 1938 to an average of 2301 cal daily in 1943. But over these same years the average body weights of the men in these factories declined only .80kg and the women workers gained an average of .45 kg.

Relevance to EPK and restrictive dieting: The surrounding paragraph says the recorded calorie figures are unreliable and points to unrecorded food purchases. This is not evidence of unchanged weight despite a verified deficit. For EPK, the lesson is to measure actual intake and expenditure before attributing a plateau to metabolic adaptation.

p. 119 | Appetite, tissue recovery and refeeding

Keys and colleagues, Volume 1, p. 119.

During the first week of uncontrolled rehabilitation (R13) the individual intakes varied from 4400 cal to 5800 cal, with the week’s average of 5219 cal per man per day. Some of the men commented that they were still hungry at the end of the very large meals, even though they were unable to ingest any more food. All of the men ate snacks between meals and in the evenings.

Relevance to EPK and restrictive dieting: Remaining hungry after very large meals distinguishes the drive to eat from immediate stomach fullness. This is relevant to EPK's behavioural account of restriction and recovery. It does not establish a single hormone mechanism or mean every episode of hunger requires unrestricted intake.

p. 127 | Appetite, tissue recovery and refeeding

Keys and colleagues, Volume 1, p. 127.

At R58 the body weight had returned close to the control value while the body fat, on the percentage basis, remained slightly elevated. In the undernutrition experiment carried out by Benedict et al. the weight increased very rapidly on the termination of the reduced diet. Six weeks later the men not only regained all the weight lost but exceeded their pre-experimental weight by 3.1 kg.

Relevance to EPK and restrictive dieting: The passage distinguishes recovering body weight from recovering body composition and also refers to another experiment. EPK should specify which recovery endpoint it means. The historical observations cannot establish that weight overshoot is inevitable or permanent after routine dieting.

p. 128 | Appetite, tissue recovery and refeeding

Keys and colleagues, Volume 1, p. 128.

they had very low resistance to eating snacks between meals. Some men had eaten a large meal; subject No. 27 commented at the end of R13 on having “an odd sensation of being full yet still hungry.”

Relevance to EPK and restrictive dieting: Feeling full yet hungry illustrates why adherence after deprivation may not be explained by motivation alone. It supports studying appetite and recent restriction history separately from meal size, without diagnosing all overeating as a starvation response.

p. 129 | Appetite, tissue recovery and refeeding

Keys and colleagues, Volume 1, p. 129.

In refeeding, especially when the caloric intake is high, fat deposits tend to increase at a faster rate than the “active” tissues.

Relevance to EPK and restrictive dieting: Faster fat than active-tissue restoration makes recovery quality relevant to EPK. It challenges equating rapid regain with complete functional recovery, but it also argues against promising that raising calories necessarily restores lean tissue preferentially or causes fat loss.

p. 177 | Appetite, tissue recovery and refeeding

Keys and colleagues, Volume 1, p. 177.

During the subsequent 8 weeks there was further increase in the body fat which brought the fat content of the body in all subjects well above the pre-starvation level. Thirty-three weeks after the end of the semi-starvation period the values for body fat showed a decrease and returned to a near normal level at 58 weeks.

Relevance to EPK and restrictive dieting: The time course described suggests that recovery-related fat overshoot may change over many months. EPK should follow longer-term outcomes rather than classify early regain as failure. This selected historical recovery pattern cannot predict each dieter's eventual body composition.

p. 179 | Appetite, tissue recovery and refeeding

Keys and colleagues, Volume 1, p. 179.

Anthropometric data support this conclusion and indicate that during rehabilitation the adipose tissue increased in size more rapidly than the muscles. In the highest caloric group (T) the circumferences of the upper arm, calf, and thigh showed an average recovery of 45, 46, and 54 per cent of the starvation decrement, whereas the abdominal circumference exhibited a recovery of 101 per cent [at week 12 of refeeding].

Relevance to EPK and restrictive dieting: Unequal changes in limb and abdominal measurements suggest that tissue recovery is not uniform. For EPK, strength and body composition would complement scale weight. Circumferences are imperfect proxies and do not alone establish visceral fat, muscle restoration or metabolic health.

p. 212 | Endocrine and reproductive function

Keys and colleagues, Volume 1, p. 212.

The evidence is quite consistent that the human thyroid atrophies during starvation.

Relevance to EPK and restrictive dieting: The thyroid observation supplies historical context for endocrine changes in severe starvation. It supports asking whether restriction compromises function, but organ atrophy in this context cannot diagnose hypothyroidism in a dieter or justify thyroid treatment.

p. 217 | Endocrine and reproductive function

Keys and colleagues, Volume 1, p. 217.

Chakrabarty, in a study of famine victims in India, found the islet cells fewer in number than normal and foamy in structure.

Relevance to EPK and restrictive dieting: The islet-cell report concerns famine pathology rather than a controlled Minnesota outcome. It may inform questions about severe undernutrition and pancreatic function, but it cannot show that ordinary calorie deficits destroy islet cells or explain all glucose intolerance.

p. 240 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 240.

Two reports from Japanese prison camps have indicated that during the periods of serious food restrictions most cases of acne disappeared. After liberation, when the food intake was restored to normal, the acne returned to its pre-starvation form. […] Ivanovsky reported that the hair grew slowly, fell out prematurely, and rapidly became gray. The growth of the nails was retarded.

Relevance to EPK and restrictive dieting: The acne account shows why improvement in one symptom during deprivation is not evidence of overall health; the hair and nail account points in the opposite direction. EPK should assess multiple functions. These reports do not isolate calories from illness or specific nutrient deficiencies.

p. 242 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 242.

The dry and scaly skin was probably an expression of an altered skin metabolism as a result of a decreased blood flow through the skin. There was indirect evidence that the blood flow through the skin had been decreased during semi-starvation.

Relevance to EPK and restrictive dieting: The proposed link between skin changes and circulation gives EPK a hypothesis about peripheral function during restriction. The source calls the evidence indirect, so reduced blood flow must not be presented as a measured universal explanation for dry skin.

p. 243 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 243.

About a third of the subjects remarked that their hair was falling out at an abnormally rapid rate.

Relevance to EPK and restrictive dieting: Reported hair loss is relevant to the broader cost of severe restriction and deserves attention beyond weight change. It remains a nonspecific symptom: nutritional deficiencies, endocrine conditions and other stressors require consideration rather than an automatic EPK classification.

p. 256 | Blood, immunity and tissue function

Keys and colleagues, Volume 1, p. 256.

The evidence is overwhelming that anemia develops during prolonged periods of caloric restriction, and the degree of anemia appears to be related to the extent of starvation […] The anemia […] is not of the iron deficiency type.

Relevance to EPK and restrictive dieting: The anaemia discussion challenges the assumption that tissue and blood function are preserved whenever weight is falling. It supports broader monitoring during restriction, but historical exclusion of iron deficiency does not establish the cause of anaemia in a current individual.

p. 258 | Blood, immunity and tissue function

Keys and colleagues, Volume 1, p. 258.

evidence indicates that a mild leucopenia with a relative lymphocytosis develops.

Relevance to EPK and restrictive dieting: Changes in white-cell patterns make immune measures a possible EPK research outcome. Counts alone do not measure immune competence or establish Proportional Danger Theory, and historical terminology should not be substituted for a modern diagnostic interpretation.

p. 267 | Blood, immunity and tissue function

Keys and colleagues, Volume 1, p. 267.

A definite leucopenia developed during semi-starvation; the leucocyte count, which was 6346 per cu. mm. of blood in the control period, dropped to 4129 at S24, a decrease of 34.9 per cent.

Relevance to EPK and restrictive dieting: The reported decline in leucocyte count offers a quantitative endpoint for investigating severe restriction. It does not by itself prove greater infection risk or an energy-to-immune causal mechanism; those require independently measured clinical and immune outcomes.

p. 291 | Metabolism and measurement

Keys and colleagues, Volume 1, p. 291.

If our calculations are approximately correct, there remains the question as to how much of the reduction in metabolism can be accounted for by the lowered body temperature of the starving organism.

Relevance to EPK and restrictive dieting: The authors' question links temperature regulation and metabolic adaptation. EPK can test those variables together, but the excerpt is a question rather than a demonstrated explanation, and lower temperature alone cannot quantify an energy shortfall.

p. 333 | Metabolism and measurement

Keys and colleagues, Volume 1, p. 333.

After 12 weeks of controlled rehabilitation on restricted food intakes, the men in the Minnesota Experiment were allowed to eat what they would. The result was a large increase in food intake, a sudden gain in weight, and an abrupt rise in B.M.R. In some cases the B.M.R. change was startling. For example, subject 126 used 153 cc. of oxygen per minute in basal rest on October 17 and 265 cc. on October 24. October 20 was the last day of restricted feeding, so the rise from 153 to 265 cc. of oxygen per minute occurred in 3 days. In this period of 3 days the subject gained 6 pounds and the basal pulse rate rose from 36 to 60.

Relevance to EPK and restrictive dieting: The rapid individual response suggests that some components of suppressed expenditure can change quickly with refeeding. This supports studying recovery dynamically. Oxygen consumption is not itself a calorie intake, and a three-day weight gain cannot be assumed to be new fat or muscle.

p. 338 | Metabolism and measurement

Keys and colleagues, Volume 1, p. 338.

In the Minnesota experiment the total rate of basal metabolism at the end of semi-starvation was almost 40 per cent less than in the control period.

Relevance to EPK and restrictive dieting: The substantial fall in total basal metabolism supports adaptation as a serious consequence of prolonged restriction. Total BMR also falls when metabolically active tissue is lost; separating tissue loss from adaptive suppression requires adjustment. It does not invalidate energy balance or define an EPK minimum.

p. 346 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 346.

Schenk (1936) claimed that daily intakes of 320 gm. of protein and 7300 cal, as observed among the Olympic athletes, are essential for hard manual work or strenuous exercise.

Relevance to EPK and restrictive dieting: Keys and colleagues immediately challenge Schenk’s claim, citing evidence that adequate carbohydrate and fat can meet the energy cost of work without the assumed increase in protein breakdown. This supports distinguishing energy requirements from exaggerated protein claims; it does not endorse a universal 7,300 kcal or 320 g protein target.

p. 359 | Metabolism and measurement

Keys and colleagues, Volume 1, p. 359.

In general, at diets of 3000 cal or more, nitrogen equilibrium can be maintained at an intake of 40 gm. of protein or less. At diets below 3000 cal, nitrogen balance is harder to obtain, even at higher nitrogen intakes.

Relevance to EPK and restrictive dieting: The passage highlights interaction between total energy and protein balance: adding protein may not fully compensate for low energy intake. That is relevant to restrictive high-protein diets, but 3,000 kcal and 40 g are historical context-dependent figures, not universal requirements.

p. 365 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 365.

spontaneous reduction of physical activity is one of the most prominent features of semi-starvation.

Relevance to EPK and restrictive dieting: Reduced spontaneous activity is directly relevant to the idea that people can compensate for restriction outside planned exercise. It challenges predictions based on a fixed expenditure assumption. EPK would still need measured activity and energy data to quantify the effect in ordinary dieting.

p. 380 | Metabolism and measurement

Keys and colleagues, Volume 1, p. 380.

A more detailed study (Evans and Strang, 1931) was performed on 5 patients with an average initial weight of 157.5 kg. They received first a maintenance diet of 2413 cal containing 69 gm. of protein, which was sufficient for nitrogen equilibrium. On the reducing diet, which contained 59 gm. of protein, 10 gm. of carbohydrate, and 7 gm. of fat, totaling 335 cal., a nitrogen loss occurred, but this was amazingly small (2 gm. of nitrogen in 5.5 weeks). The addition of only 20 gm. of carbohydrate (increasing the total caloric intake to 445 cal. per day) brought these patients into nitrogen equilibrium at an excretion level of 0.15 gm. of nitrogen per kg. of “ideal” weight. Nitrogen equilibrium was maintained in one patient for 260 days, during which time there was a total weight loss of 67.1 kg.

Relevance to EPK and restrictive dieting: These observations in a few patients with severe obesity challenge extrapolation from lean semi-starved volunteers to everyone. EPK should address body stores and study context. Nitrogen equilibrium is a narrow outcome and cannot establish that a 335-445 kcal diet is safe or broadly effective.

p. 381 | Metabolism and measurement

Keys and colleagues, Volume 1, p. 381.

The loss of fat produces no untoward symptoms such as accompany the weight loss of normal subjects during semi-starvation. Furthermore, no decrease of the B.M.R. below normal standards has been reported in obese patients during the period of weight reduction.

Relevance to EPK and restrictive dieting: This historical generalisation should be presented as a counterargument, not omitted. It does not establish today's claim that people with obesity never experience metabolic adaptation or symptoms during weight loss; the standards, sample and available measurements require scrutiny.

p. 518 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 518.

The behavior of calcium is more difficult to interpret since it tends to be excreted throughout a starvation period at a rate which may be of the order of 10 times that calculated on the basis of the body tissue lost.

Relevance to EPK and restrictive dieting: Calcium loss raises questions about mineral balance and skeletal consequences that weight alone will miss. It broadens EPK's outcomes, while leaving uncertain how much is due to energy restriction, nutrient intake or other physiology. It does not specify a calcium supplement treatment.

p. 545 | Metabolism and measurement

Keys and colleagues, Volume 1, p. 545.

To test this, he reduced his own food intake by one half for 5 days and then for an additional 3 days went completely without food. The increase in his blood sugar level following glucose was much greater after the starvation period than before. Sevringhaus observed the same thing after 2 days of starvation.

Relevance to EPK and restrictive dieting: The fasting observations suggest that a glucose challenge can depend on recent nutritional exposure. EPK studies should standardise pre-test intake. A transient response after fasting is not proof of chronic diabetes or a general argument that calorie restriction permanently impairs glucose regulation.

p. 546 | Metabolism and measurement

Keys and colleagues, Volume 1, p. 546.

Because of the high fat diet of the Eskimos, Heinbecker (1928) studied their glucose tolerance curves both before and after 82 hours of fasting. The basal blood sugar during the starvation decreased from 110-120 mg. per 100 cc. to about 80 mg. The glucose tolerances after the ingestion of about 2 gm. of glucose per kg. of body weight were normal in the control period. After 3.5 days of starvation the glucose tolerance showed blood sugar levels close to 300 mg. per 100cc. These returned only very slowly to normal.

Relevance to EPK and restrictive dieting: This historical account concerns a particular population and fasting exposure. It supports caution when comparing glucose tests across dietary states, not a universal conclusion about low-carbohydrate diets, ethnicity or permanent metabolic damage.

p. 578 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 578.

Along with the increased water consumption, starving persons tend to have a marked salt hunger and will consume several times the normal quota of salt if it is available.

Relevance to EPK and restrictive dieting: Salt seeking and greater drinking suggest that deprivation can alter more than appetite for calories. EPK should distinguish hydration and electrolyte issues from hunger. The observation does not justify unrestricted salt or fluid intake or diagnose an individual electrolyte deficit.

p. 579 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 579.

In some animal forms, at least, chronic undernutrition prolongs the natural life span. It has been suggested that the natural life span is fixed, not in time, but in terms of total metabolism or some function of the rate of living. But in man severe undernutrition makes him look, feel, and act prematurely old. There are also changes in basal metabolism and in sexual function which resemble those produced by age. What is the long-range effect on physiological age and on longevity of the individual?

Relevance to EPK and restrictive dieting: Looking and feeling older during severe undernutrition fits EPK's interest in functional capacity. The passage explicitly contrasts this with animal longevity findings and asks an unresolved question. It cannot establish faster human biological ageing or shorter lifespan from all restriction.

p. 581 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 581.

the starving man is weak and cold, both physiologically and subjectively, and his behavior bears this out. On the other hand, his behavior is often misleading. He acts dull and insensitive; he looks and behaves as though he were unaware of or incapable of feeling many of the ordinary stimuli of sound, sight, or touch.

Relevance to EPK and restrictive dieting: Outward dullness may conceal substantial physiological discomfort. This supports interpreting reduced engagement in context rather than as character failure. EPK must still investigate alternative medical or psychological explanations rather than attribute every such behaviour to intake.

p. 585 | Metabolism and measurement

Keys and colleagues, Volume 1, p. 585.

in a number of cases among the women the decline in metabolic rate was disproportionately great; only small losses in body weight sufficed to produce large declines in metabolism among these women.

Relevance to EPK and restrictive dieting: The reported mismatch between weight change and metabolic change is especially relevant to EPK's criticism of weight-only monitoring. Its magnitude and applicability require the underlying women's study; the excerpt cannot define a sex-specific threshold or show what caused each decline.

p. 607 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 607.

In any list of the cardinal signs and symptoms of severe undernutrition there will be several items which pertain directly or indirectly to the heart and circulation. These include bradycardia, hypotension, a lowering of the skin temperature, and frequently vertigo and slight cyanosis. […] Rates as low as 30 to 40 beats per minute were often observed in bed rest […]

Relevance to EPK and restrictive dieting: Slow pulse, hypotension and circulatory symptoms can accompany severe undernutrition. They make safety and function relevant to assessing a restrictive diet. A slow pulse alone is not proof of starvation and must not be confused automatically with athletic conditioning.

p. 612 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 612.

The complaint of coldness is very common among undernourished people, and the skin has often been reported as cold to the touch. This might, of course, be attributed to the general reduction in body temperature which is usually observed. Leyton found the average 4 p.m. temperature in 100 men to be 35.7 C (96.3F). But the suggestion that the peripheral circulation is diminished is also supported by the pallor of the skin which is out of proportion to the slight to moderate anemia usually present.

Relevance to EPK and restrictive dieting: The passage links coldness to both body temperature and possible peripheral circulation changes. EPK could measure these separately. It does not establish that all cold hands or low temperatures result from dieting, nor that calorie increases alone will resolve them.

p. 613 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 613.

In 40 persons who received subcutaneous injections of 1 mg. of adrenalin there was extraordinarily little response to the drug.

Relevance to EPK and restrictive dieting: A reduced response to injected adrenaline suggests that stress-hormone effects can depend on physiological state. This complicates a simple claim that under-fuelling always produces excessive adrenaline symptoms; it does not establish a modern condition of adrenal exhaustion.

p. 620 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 620.

Before the war patients with hypertension accounted for about 10 per cent of all admissions to the Therapeutic Clinic of the Pavlov First Medical Institute. This proportion dropped during the period of semi-starvation and increased progressively thereafter through 1943. In January 1943 patients with hypertension constituted 20 per cent of the admissions, and by June of that year the figure had risen to 60 per cent.

Relevance to EPK and restrictive dieting: Changes in the share of hospital admissions for hypertension may reflect many changes in illness and access as well as nutrition. For EPK this is contextual historical evidence, not an estimate of population risk or proof that refeeding causes hypertension.

p. 621 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 621.

Obviously, many factors may have been involved in this epidemic of hypertension in Leningrad, but the outstanding peculiarity of the period was severe semi-starvation for 6 months, followed by refeeding, with the necessity for hard work at all times.

Relevance to EPK and restrictive dieting: The authors themselves acknowledge multiple possible causes in a setting combining deprivation, refeeding and hard work. EPK should similarly separate exposures. The lesson for restrictive dieting is contextual interpretation, not a universal prediction of hypertension after eating more.

p. 638 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 638.

An almost universal complaint among the subjects in the Minnesota Experiment was some degree of faintness immediately after standing up from a chair or bed.

Relevance to EPK and restrictive dieting: Postural faintness is a functional outcome that a weight-loss chart would miss. It supports taking symptoms during restriction seriously, while recognising dehydration, medication and other conditions as alternatives. It is not a diagnostic EPK test.

p. 659 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 659.

The decrease of the QRS and T wave amplitudes is probably the most remarkable change in the electrocardiogram in starvation. It may be due to the following factors: 1) Decrease of heart size. The decrease of the amplitudes was accompanied by a very marked decrease of the heart size as determined by teleroentgenograms […] 2) Myocardial degeneration […] 3) Rotation of the heart around a transversal axis… 4) Decreased metabolic rate… 5) Fluid accumulation in chest or pericardium […]

Relevance to EPK and restrictive dieting: The list presents possible explanations for cardiac electrical changes, including heart size, metabolism and fluid. It motivates cardiac outcomes in severe-undernutrition research, but none of the listed mechanisms should be presented as proven by this quotation or generalised to ordinary dieting.

p. 664 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 664.

Polyuria and nocturia have been reported by practically every observer of starvation conditions.

Relevance to EPK and restrictive dieting: Nocturia offers another possible route to disrupted sleep during severe starvation, distinct from hunger or low glucose. EPK sleep studies should ask about waking to urinate. The historical statement does not show that a current person's nocturia is caused by low intake.

p. 740 | Physical function and clinical context

Keys and colleagues, Volume 1, p. 740.

The important question of what part of the respiratory-cardiovascular system was principally responsible for the poor delivery of oxygen to the working muscles cannot be answered directly. The low concentration of hemoglobin at the end of the semi-starvation period undoubtedly contributed to the poor muscular oxygen supply.

Relevance to EPK and restrictive dieting: Reduced haemoglobin could limit oxygen delivery and performance during severe restriction. This is relevant to EPK's functional emphasis, but the authors explicitly state that the principal limitation was unresolved. It does not demonstrate a primary mitochondrial defect.

p. 753 | Endocrine and reproductive function

Keys and colleagues, Volume 1, p. 753.

Jaworski found no changes in the vagina in starved Polish women but concluded there was considerable reduction in the size of the ovaries. […] The majority of investigators have agreed that the testes are reduced in size in semi-starvation, the average percentage change being rather similar to that of the body as a whole.

Relevance to EPK and restrictive dieting: Reported reproductive-organ changes broaden the consequences of severe deprivation beyond body fat. EPK can use them as historical context, but they derive from different populations and measures, not a direct basis for predicting organ changes in routine dieters.

p. 755 | Endocrine and reproductive function

Keys and colleagues, Volume 1, p. 755.

The excretion of total, neutral 17-ketosteroids for the 5 men studied after 8 weeks of semi-starvation averaged 7.9 mg. per 24 hours. This is considerably below the mean excretion level of 11.3 mg. observed for normal young men […]

Relevance to EPK and restrictive dieting: The steroid-excretion finding is relevant to endocrine responses to undernutrition. Urinary 17-ketosteroids are not interchangeable with a modern serum testosterone result; the small historical sample cannot establish a testosterone threshold or a treatment indication.

p. 762 | Endocrine and reproductive function

Keys and colleagues, Volume 1, p. 762.

Table 359 summary: in the four-subject Minnesota subgroup, mean sperm longevity (hours motile) was 4.8 hours at S24 and 25.5 hours at R20.

Relevance to EPK and restrictive dieting: This four-person subgroup illustrates a change in reproductive function during recovery. It supports measuring function as well as body mass in EPK research, but the subgroup must not be confused with the larger sample discussed in the surrounding text or treated as a population fertility estimate.

p. 777 | Physical function and clinical context

Keys and colleagues, Volume 2, p. 777.

The official daily ration in the Auschwitz concentration camp was one liter of watery soup, 250 gm. of bread, and about 25 gm. of margarine or sausage or imitation honey, providing an estimated maximum intake of 1000 cal. per day.

Relevance to EPK and restrictive dieting: This passage documents extreme deprivation under atrocity conditions. It belongs in the historical record, with its context intact. It should not be used to equate a voluntary weight-management diet with concentration-camp persecution or to derive routine nutritional targets.

p. 797 | Physical function and clinical context

Keys and colleagues, Volume 2, p. 797.

Constipation was mentioned as one of the semi-starvation symptoms, bowel movements being reduced from once a day to one or two times a week.

Relevance to EPK and restrictive dieting: Less frequent bowel movements can accompany reduced food intake and severe undernutrition. EPK can include digestive function, but stool frequency also depends on food volume, fibre and other factors. This does not establish that every restrictive diet causes a gastrointestinal disorder.

p. 833 | Behaviour, cognition and recovery

Keys and colleagues, Volume 2, p. 833.

Food in all its ramifications became the principal topic of conversation, reading, and daydreams for almost all Minnesota subjects. When they read books or attended movies, they were much impressed by the frequency with which food and eating were mentioned. Cookbooks, menus, and information bulletins on food production became intensely interesting to many of the men who previously had had little or no interest in dietetics or agriculture.

Relevance to EPK and restrictive dieting: Food preoccupation is directly relevant to the argument that persistent deprivation can reshape attention and behaviour. It supports studying diet burden rather than blaming a person for thinking about food. It cannot show that all food interest or cravings indicate energy deficiency.

p. 834 | Behaviour, cognition and recovery

Keys and colleagues, Volume 2, p. 834.

A few planned to become cooks. Halfway through starvation 13 out of 34 men mentioned cooking among their plans following completion of the experiment. The men who took part in Greely’s expedition of 1881 and suffered from prolonged semi-starvation in the Arctic had similar ideas. In the diary entry for November 24, 1883, Brainard wrote: […] Fredericks is going to run a saloon in Minneapolis. Long wants to open a restaurant at Ann Arbor. Jewell thinks he would like to run the grocery in Ralston’s colony. […]

Relevance to EPK and restrictive dieting: Changing occupational interests towards food illustrates how deprivation may reorganise priorities. For EPK this is descriptive behavioural context, not a diagnostic symptom: the passage also draws on another expedition and cannot quantify effects of moderate restriction.

p. 827-828 | Behaviour, cognition and recovery

Keys and colleagues, Volume 2, p. 827-828.

The subjects reported that their nails grew more slowly and that their hair was falling out in large amounts. Shaving was necessary less frequently. The men noted, particularly in shaving, that cuts and wounds bled less than normally and were slower to heal. Physical ability to laugh heartily, sneeze, or blush was reduced or absent during the later stages of semi-starvation. Muscle cramps and particularly muscle soreness were frequently reported. The jarring of knee joints, especially when walking on hard pavements, was an annoyance to some. Pigmentation, thinning, and roughening of the skin occurred. Changes in the sensitivity of the skin, paresthetic and hypesthetic in character, were observed in only three cases, but there were many complaints that the extremities “went to sleep.” Tolerance to heat was increased; for example, subjects could hold hot plates without discomfort. They asked that their food, coffee, and tea be served unusually hot. Conversely, cold temperatures were poorly tolerated. Complaints of being cold or of having cold hands and feet were frequent and persistent. In hot summer weather many of the subjects slept under heavy blankets and wore extra clothing during the day. Vertigo, giddiness, and momentary blackouts were experienced on rising from lying or sitting positions by almost all the subjects during the first months and by some subjects throughout the semi-starvation period […] Objective tests revealed no impairment of visual acuity but many subjects complained of transient visual disturbances such as inability to focus, eye-aches, and “spots” before their eyes. Standard measurements of hearing showed a slight but consistent increase in auditory acuity during the period of semi-starvation. It is difficult to determine whether the frequent complaints that ordinary sounds and noises were disturbing and annoying had a direct physiological basis in the “improved” auditory sensitivity or were primarily signs of an increased irritability. Sensations of ringing in the head were reported. Except for hunger pains and some decrease in the frequency of bowel movements, gastrointestinal symptoms were rare. The marked decreases in pulse rate and basal metabolism may be regarded as critical indicators of a lowering of speed in the automatic functions of the body […] The marked reduction in strength and endurance was paralleled by a general curtailment of self-initiated, spontaneous activities […] They described their increasing weakness, loss of ambition, narrowing of interests, depression, irritability, and loss of libido as a pattern characteristic of “growing old.” During rehabilitation the recovery from dizziness, apathy, and lethargy was most rapid. Tiredness, loss of sex drive, and weakness were slow to improve. Although visible edema tended to disappear, in some men there was little change or even an increase in edema. Cramps, vague aches and pains, and paresthesias were unrelieved for some time. Some of the men had new complaints such as flatus, distention, belching, and stomach-ache. Those subjects who gained the most weight became concerned about their increasing sluggishness, general flabbiness, and the tendency of fat to accumulate in the abdomen and buttocks. At the end of 3 months of rehabilitation, even in those subjects who were maintained on the highest caloric intake the over-all physical condition was considerably inferior to the pre-starvation status. Later reports from the subjects indicated that it was not until after an additional 3 months of “normal” living and supernormal eating that their physical capacity approached pre-experimental levels.

Relevance to EPK and restrictive dieting: This wide-ranging account supports evaluating recovery through strength, engagement, temperature tolerance and sexual function rather than weight alone. It also reports symptoms persisting into rehabilitation, challenging the promise of an immediate reset after eating more. Individual symptoms remain nonspecific, and subjective complaints must be distinguished from objective test results.

p. 835 | Behaviour, cognition and recovery

Keys and colleagues, Volume 2, p. 835.

Satisfaction was also obtained from consumption of coffee and tea, both of which were used in large quantities, presumably for their pharmacological as well as their filling and warming effects. It was generally reported that coffee and tea provide a “lift.” Because some of the men increased their consumption to 15 or more cups daily, it became necessary to limit all subjects to a maximum of 9 cups per day […] Heavy gum chewers would take 2 or 3 sticks at a time, chew them until the sweet taste was gone, discard them, and then replace them with fresh sticks in chain fashion. One of the men chewed up to 40 packages of gum per day and developed a sore mouth from such continuous exercise. Thereafter, the use of gum was restricted to 2 packages a day.

Relevance to EPK and restrictive dieting: The coffee, tea and gum accounts suggest compensatory strategies for warmth, stimulation and appetite management. EPK could record these behaviours alongside sleep and hunger. They do not prove that caffeine use causes, diagnoses or cures under-fuelling; caffeine itself could confound sleep findings.

p. 836-837 | Behaviour, cognition and recovery

Keys and colleagues, Volume 2, p. 836-837.

Social initiative especially, and sociability in general, underwent a remarkable change. The men became reluctant to plan activities, to make decisions, and to participate in group activities. Their earlier active interest in having a voice in the making of policies and rules for the conduct of the non-scientific aspects of the experiment dwindled. They spent more and more time alone.

Relevance to EPK and restrictive dieting: Reduced social initiative supports considering withdrawal as a possible consequence of severe physiological strain. This informs an anti-blame account of diet adherence, but does not establish that low intake explains all social avoidance or removes the need to assess mood and social context.

p. 839 | Behaviour, cognition and recovery

Keys and colleagues, Volume 2, p. 839.

Sexual feeling and expression declined in the Minnesota Experiment until by the end of the semi-starvation period they were virtually extinguished in all but a few subjects. The diminution of the strength of the sex drive was so dramatic that the subjects were struck by the change and used colorful language to describe it.

Relevance to EPK and restrictive dieting: The decline in sexual interest makes wellbeing and reproductive function relevant outcomes of restriction. For EPK, those outcomes deserve attention alongside weight. Libido alone neither measures testosterone nor identifies energy deficiency, and recovery cannot be promised from a fixed intake target.

p. 846 | Behaviour, cognition and recovery

Keys and colleagues, Volume 2, p. 846.

During […] (R13), of 17 men who made a reference to food cravings, 7 had no specific cravings. The remaining 10 men had a predilection, in about equal proportion, for sweets (ice cream and pastries) and dairy products.

Relevance to EPK and restrictive dieting: The variation in food cravings matters: the passage includes men reporting no specific cravings. EPK should measure individual responses rather than predict a universal drive for sweets. The observations do not establish that a particular food preference reflects a specific deficiency.

pp. 852-853 | Behaviour, cognition and recovery

Keys and colleagues, Volume 2, pp. 852-853.

I’m cold […] I’m weak […] And now I have edema […] Social graces, interests, spontaneous activity and responsibility take second place to concerns of food […] I don’t like to sit near guests, for then it is necessary to entertain and talk with them […] I am one of about three or four who still go out with girls. I fell in love with a girl during the control period but I see her only occasionally now. It’s almost too much trouble to see her even if she visits me at the Lab. It requires effort to hold her hand. Entertainment must be tame. If we see a show, the most interesting part of it is contained in scenes where people are eating. I couldn’t laugh at the funniest picture in the world, and love scenes are completely dull […] Now, eight months after the end of starvation, I am fat and healthy although my muscles have not yet returned to their former tone.

Relevance to EPK and restrictive dieting: This individual's account makes the subjective and social burden of deprivation tangible and distinguishes weight restoration from muscular recovery. It can illustrate the EPK rationale, but one testimony cannot estimate prevalence, establish mechanisms or represent all dieters.

p. 859 | Behaviour, cognition and recovery

Keys and colleagues, Volume 2, p. 859.

As semi-starvation progressed in the Minnesota Experiment, complaints of inability to concentrate for any period of time and of difficulty in developing thoughts became numerous.

Relevance to EPK and restrictive dieting: Concentration complaints give EPK a reason to study attention during under-fuelling. They are subjective reports rather than proof of objective cognitive impairment or permanent brain injury. A stronger test would pair symptoms with validated cognitive measures and control for sleep.

pp. 864-865 | Behaviour, cognition and recovery

Keys and colleagues, Volume 2, pp. 864-865.

There was a statistically highly significant increase in scores on the scales of social introversion, depression, and cycloid tendencies (emotional instability) and a decrease in those on the scales of ascendancy (social leadership), self-confidence, and freedom from nervous tenseness.

Relevance to EPK and restrictive dieting: Changes in psychological scale scores suggest that deprivation can affect emotional and social functioning, strengthening a contextual account of adherence. Historical scales and labels should not be equated with modern diagnoses, and no single score demonstrates an EPK state.

p. 880 | Behaviour, cognition and recovery

Keys and colleagues, Volume 2, p. 880.

In four subjects the reaction to the semi-starvation regimen took the form of a “character neurosis,” the men being unable to stay on the semi-starvation diet; in two of these subjects the response to the stress was particularly violent and bordered on a psychosis. One man developed a hysteroid reaction that led to self-mutilation.

Relevance to EPK and restrictive dieting: These severe individual reactions illustrate why prolonged semi-starvation is not an ethically appropriate model to reproduce casually. They challenge moralising about inability to maintain restriction, but outdated psychiatric labels and exceptional events cannot be generalised to ordinary dieting or used to explain all mental illness.

p. 968 | Physical function and clinical context

Keys and colleagues, Volume 2, p. 968.

At equal degrees of weight loss, anorexia nervosa patients are similar to famine victims in numerous respects: the pulse is slow, averaging about 50; the blood pressure is low, rarely above 120 mm. Hg. Systolic; amenorrhea is almost universal. The basal metabolic rate is reduced to an average of −30 per cent […], but values down to −73 per cent have been obtained […] The body temperature is low […]

Relevance to EPK and restrictive dieting: Similarities in physical signs across undernourished groups support recognising starvation physiology regardless of how it began. They do not show that anorexia nervosa is simply a calorie-deficit condition; its psychiatric and medical dimensions require independent assessment.

p. 978 | Historical disease and developmental context

Keys and colleagues, Volume 2, p. 978.

It has been repeatedly demonstrated in many species of laboratory and domestic animals that certain types of congenital malformations frequently appear in the offspring as a result of maternal nutrition deficiencies.

Relevance to EPK and restrictive dieting: Animal developmental findings make maternal nutrition a relevant research context for EPK. The passage concerns nutritional deficiencies broadly, not an isolated human calorie-deficit experiment. It cannot establish a human calorie target or attribute a congenital condition to maternal behaviour.

p. 979 | Historical disease and developmental context

Keys and colleagues, Volume 2, p. 979.

Burke et al. observed a correlation between the adequacy of the maternal diet and the pediatric rating of the offspring. Almost all the cases of congenital defects were found among the infants born to mothers whose diet during pregnancy had been severely inadequate.

Relevance to EPK and restrictive dieting: The historical maternal-diet association raises a developmental question but cannot isolate calories from micronutrients, illness or social conditions. EPK should avoid causal or blame-based claims and use modern pregnancy-specific evidence before making recommendations.

p. 1003 | Historical disease and developmental context

Keys and colleagues, Volume 2, p. 1003.

There is reason to believe that the epithelial lining of the gastrointestinal tract becomes more permeable to microorganisms in severe undernutrition. The morphological changes in the intestinal tract would strongly suggest this.

Relevance to EPK and restrictive dieting: The proposed intestinal-barrier mechanism is relevant to the immune paper as a hypothesis about severe undernutrition. The passage is explicitly suggestive, not direct proof of permeability or infection causation. It cannot validate Proportional Danger Theory or diagnose a gut-barrier problem in dieters.

p. 1041 | Historical disease and developmental context

Keys and colleagues, Volume 2, p. 1041.

There is a fundamental connection between obesity and diabetes. Kisch found that 50 per cent of all the obese persons he examined exhibited some degree of glycosuria.

Relevance to EPK and restrictive dieting: The historical glycosuria observation concerns obesity and glucose handling, not direct evidence that restriction causes disease. It is a counterweight to treating higher intake as universally beneficial. Glycosuria is not itself sufficient to reconstruct diabetes prevalence from this excerpt.

p. 1042 | Historical disease and developmental context

Keys and colleagues, Volume 2, p. 1042.

It is proper to suggest that obesity, which first may only resemble diabetes superficially in a tendency to produce glycosuria and abnormal sugar tolerance tests, may slowly produce irreversible changes, even though weight reduction may correct the glycosuria temporarily. Allen supports the general clinical impression that in obese persons in whom glycosuria disappears during dietary reduction, true diabetes is very apt to appear later. […] Mahaux and others have concluded that there was a real increase in diabetes morbidity in Germany immediately following World War I, and that this was associated with an increased food intake following a period of undernutrition.

Relevance to EPK and restrictive dieting: These passages raise questions about underlying disease and the timing of restriction and refeeding, but they report historical impressions and associations. They cannot show that dieting causes later diabetes, that weight reduction is futile, or that refeeding alone explains morbidity.

p. 1048 | Historical disease and developmental context

Keys and colleagues, Volume 2, p. 1048.

For many years it has been generally believed that yeast may be beneficial to diabetics. The insulin requirement of depancreatized dogs is increased on a diet definitely deficient in B vitamins, and this extra defect may be corrected either with yeast or with a mixture of B vitamins.

Relevance to EPK and restrictive dieting: The dog experiment concerns B-vitamin deficiency superimposed on absent pancreatic function. It highlights the need to distinguish energy sufficiency from nutrient sufficiency in EPK. It is not evidence that yeast, B vitamins or more calories can replace diabetes treatment.

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.