Studies on the Destruction of Vitamin C in the Boiling of Milk

Abstract

An experiment has been reported on the destruction of Vitamin C in milk by boiling, based upon the development of experimental scurvy in guinea pigs. The data do not permit the mathematical expression of results with a precision of a few per cent. They do, however, indicate that there was no detectable difference in the destruction of vitamin C between boiling milk in aluminum and boiling in glass. The loss of vitamin is placed at about 20 per cent. If glass may be taken as an inert standard, it is certain that no detectable catalysis by the aluminum is evident; rather, as was to be expected, the effect is referable to the cooking per se. The data are also of interest in that our winter milk was approximately if not actually of the same potency as the best summer milk obtained from cows on pasture(9). If a pint of ordinary city milk is required to insure the protection of infants from scurvy, and if such milk protects guinea pigs when fed at the 80 to 100 cc. level (4), it can be said that the milk used by us was at least twice as potent, even in the boiled state, as ordinary city delivered milk in New York City. We would estimate that 200 cc. daily of our raw milk less than 12 hours old would probably suffice for the infant, but of course we would recommend more for the sake of safety. Dutcher(27) has had the general experience that 30 cc. of milk will protect 250 gram guinea pigs, and his observation is to be accounted for by the fact that poorly produced and handled milk is an unknown quantity at Agricultural Experiment Stations and Colleges in the United States. The data are of importance in that the destruction was of the order of only 20 per cent, which is in complete conformity with the historical fact that boiled milk has adequately served in the past as the sole or chief source of antiscorbutic vitamin for some artificially fed infants. However, this historical fact applies to less strenuous circumstances than operate in a modern city; and therefore, by elimination, the problem is to obtain potent milk to begin with. The data are also of considerable interest in emphasizing the necessity of some rigorous definition of the term “protective dose” at least for the basis of assaying, even if not for computing optimum allowances for the animal. The data fall in line with the experience of pharmacologists who recognize a latitude between the least dose producing a given degree of an effect and a dose producing it in practically all cases(26). This therefore brings into an assay of this type a statistical aspect, regardless of whether arithmetical computations employed by statisticians are or are not performed. It might be imagined that discrepancies between the value of equal-size doses would occur in an animal from day to day, much as with single doses of a medicament given to different animals at the same time. It would appear plausible under these circumstances that unequal effects from equal doses of vitamin repeated daily would reach an average in the long run, but the theory unfortunately does not work out in practice, as evidenced by remissions of disease in the same animal, or by a partial incidence of disease in animals at a single dose level. When this is realized, it is only one step further to conclude that the agreement in a small amount of data is certainly not indicative of high accuracy, even in assays involving repeated daily feeding. The growth curves and food intake data were of no help in drawing our conclusions. This is not to be interpreted as a general prediction for the future. Potent milk is absolutely essential to the success of an experiment of this sort. While the pipette feeding of milk to guinea pigs does not appear to be the best biological practice, and apparently influences appetite adversely, we believe that it is superior to cup feeding, which introduces spillage, aging, cream consumption, and varying daily doses.