Introduction. 

With the increasing human activities in circumpolar areas in recent years, the problem of an adequate and physiologically suitable ration with high caloric density has been the subject for much dispute among physiologists and nutritionists. Although a high protein diet has been successfully used by Eskimos and arctic travellers for generations, both as trail diet and emergency rations, there has been a tendency in recent years to emphasize the desirability of an all-carbohydrate ration ( Mellinger, 1948, Roth, 1948, Dyme, 1950). One of the main arguments in favor of the carbohydrate ration has been its antiketogenic effect, assuming harmful effect of slight ketosis even during the short periods of survival in question under arctic conditions. 

With reference to these questions a study was designed, the purpose of which was to consider a high-carbohydrate versus a high-protein diet, with reference to physiological adequacy and an evaluation of the physiological and clinical significance of ketosis under strenuous arctic field conditions during midwinter in Alaska. 

4. Discussion. 

It is clearly realized that the determination of ketone bodies is subject to considerable inaccuracy. However, the reported data indicate that under the conditions of the test the excretion of urinary acetone in the meat group never exceeded 1 g per day. The highest figure recorded was 8 66 mg which is a very small amount. 

According to Peters and Van Slyke (1946), ketones are regularly found in the urine of healthy persons leading a normal life. Van Slyke found as much as 280 mg ketones per 1000 ml urine. Others have reported figures between 7 and 125 mg daily. 

Under conditions such as total starvation when all energy is derived from protein and fat, the production of ketone bodies by the liver is accelerated and the excretion of ketones in the urine increases. In normal adults the appearance of gross ketonuria, according to Peters and Van Slyke (1946), does not reach its height until 3 to 5 days of the fast have elapsed. As starvation proceeds ketosis gradually diminishes. He states that in the normal male, ketosis of starvation does not reach serious proportions because sufficient carbohydrate is derived from protein and oxidized, and the levels of blood ketones are not high enough to tax severely the mechanism for the preservation of acid-base equilibrium. In one subject about 6 g of ii-hydroxybutyric acid were excreted daily in the urine for the last two weeks of a 31-day fast. In diabetic acidosis ketonuria may reach values 10 times higher than this. 

Compared with these figures the amounts of acetones excreted in the urine in our subjects on the meat ration are insignificant, and it appears that this slight ketonuria observed under these conditions for the periods considered likely as the duration of a survival situation would have no appreciably harmful effect. 

Exercise greatly increases the ketosis, and a 10-mile walk in the morning without breakfast will produce distinct ketonuria in a healthy person who otherwise is living on a normal diet (Courtice and Douglas, 1936). 

A number of evidences indicate a mechanism of adaptation to ketosis. In our Eskimo studies it is observed that the degree of ketonuria is less than what is normally observed in Whites on a similar diet. On the other hand, an Eskimo soldier who had lived for several months on the normal Army mess rations excreted the same amounts of acetone as the normal white soldiers when given a "ketogenic" diet. In the subjects studied by McClellan and DuBois (1930) the ketonuria diminished after several months on a carbohydrate-free diet. 

Deuel and Oulick (1932) have demonstrated that ketosis develops more rapidly and attains greater intensity in women than in men. 

It has been repeatedly observed that ketosis frequently occurs under strenuous field conditions regardless of the diet, and Sargent and Consolazio (1951) showed that the ketosis is reduced when the same subject undergoes repeated field tests, indicating some evidence of adaptation. 

In an Arctic bivouac at Fort Churchill the approximate caloric expenditure was 4000 calories per day. The caloric intake was about 3600. Under these conditions all the men showed trace quantities of urinary ketones almost every day, starting on the third day in the bivouac (Molnar et al., 1942). 

Of the great variety of physical fitness tests (Cureton, 1947), the Treadmill Test was selected for practical reasons. It should be emphasized, however, that physical fitness is exceedingly difficult to evaluate, not only because the meaning of physical fitness is far from clear, but also because the result of the test is greatly dependent upon a number of factors beyond the control of the observer. 

In all cases we observed an improvement at the end of the field phase, most marked in the carbohydrate group, associated with approximately 10-pound weight loss (7.5 %). 

It should be noted that the subjects had been living on a caloric deficit of the order of 2000 calories a day, and performing daily route marches of 10 miles. 

In the case of untrained personnel in poor physical condition, one would expect an improvement in physical fitness during the field phase. Our subjects, however, were all well trained and in excellent physical condition at the onset of the experiment. The factor of physical training therefore can hardly explain the difference in the physical fitness scores. 

On the other hand, it appears that the weight loss may be the most important factor in explaining the observed difference. The subjects started off probably slightly overweight and the loss of 7 per cent of their body weight would tend to increase their physical performance, since there is less weight to carry during the exercise. This is in conformity with general experience under similar conditions. It is observed that the carbohydrate group, which had the greatest weight loss, also showed the greatest improvement of physical fitness scores. The purpose of the experimental phase was to study the effect of the experimental diet on various physiological functions as compared with the levels during the normal conditions in the standardization phase. The results indicate the following effect: Both in the carbohydrate group and in the meat group, there was an increase of the physical fitness scores, most pronounced in the carbohydrate group. The basal heat production was 13 per cent higher at the end of the experimental phase than during the standardization phase in the meat group, while a reduction of 7 per cent occurred in the carbohydrate group. This difference is probably due to the specific dynamic action of protein. During the experimental phase the meat group consumed 300 ml more fluid per day than the carbohydrate group. While all subjects in the meat group were in a positive nitrogen balance, the subjects in the carbohydrate group showed a negative balance of 6.3 g on an average. Ketonuria occurred in all meat subjects and in three of the carbohydrate subjects. 

During the field phase the factor of climatic stress was added to the experimental conditions, and the following results were obtained: 

No significant difference was observed in the physical performance of the subject on the meat ration, the carbohydrate ration, or on the meat-and-carbohydrate ration during the actual field phase. The physical fitness scores were improved in all three groups at the end of the field phase, and this improvement was greatest in the carbohydrate group which also had the greatest weight loss. The psychiatric evaluation revealed no distinct differences between the three groups. There was no significant deterioration in morale, but an increase in carelessness, irritability and desire to sleep which occurred in all three groups. The weight loss was 7.0 per cent in the meat group, 7.5 per cent in the carbohydrate group, and 6 per cent in the group receiving both meat and carbohydrate. There was an increase in the basal heat production of 9 per cent in the meat group and 7 per cent in the meat-carbohydrate group, while the carbohydrate group showed a reduction of 7 per cent in the BMR. The water consumption was 1500 ml in the meat group, 850 ml in the carbohydrate group, and 1200 ml in the meat-carbohydrate group. All three groups showed negative nitrogen balance, which was most pronounced in the carbohydrate group, where it was approximately 7 g, as against approximately 2 g in the meat group. Ketonuria occurred in all three groups, most pronounced in the meat group. 

On the basis of these findings, and in view of the fact that water supply, as a rule, does not present any problem in the Arctic, it may be concluded that the carbohydrate ration offered no significant advantage under conditions of arctic survival as stimulated in the present study. In terms of heat production and nitrogen balance, the high meat ration is preferable. It is evident from this study that under survival conditions, which necessitate caloric expenditure, between 2500 and 3000 calories per man per day, including travel of approximately 10 miles a day, 1000 calories per man per day is sufficient for a period of at least 10 days. 

It would therefore seem logical that survival rations developed for arctic use should consist of protein, fat, and carbohydrate in proportions which would serve to utilize the specific dynamic action of a high protein diet, the high caloric density of fat, and the physiological advantages of carbohydrates. Protein-fat rations with high caloric density such as various types of pemmican, have already been successfully used for more than half a century by arctic travellers. 

It would appear advisable to base future arctic survival rations on the principle of a high meat-fat ration as the main meal of the day prepared in the evening, and an all-carbohydrate component of the ration to be consumed in the middle of the day while on the trail. 

5. Summary and Conclusions. 

In a series of laboratory experiments followed by field experiments under strenuous arctic conditions, the physiological adequacy of low caloric arctic rations have been studied in groups of normal men under conditions which necessitate travel under various arctic conditions. The rations studied contained approximately 1000 calories per man per day and consisted of an all-carbohydrate ration and a high protein-fat ration. On the basis of the presented data it may be concluded that from a physiological standpoint the all-carbohydrate ration offered no significant advantage over the high meat ration under the conditions of the study.