has been carried out, the laboratory rat being used, in the main, as the experimental animal.

Little experimental work has been done in fat nutrition, however, among human subjects. For that reason, the present study was decided on. To insure valid results, examinations of the 267 children (white) included in this study were made for a period of 2 years.

1

As early as 1925 Holmes, 1 in studies carried out on human subjects, found that margarine was from 93 to 97 percent digestible. Bunker,' in 1927, stated:

"Beef fat excreted in the milk of the cow is no different in its origin from beef fat which is retained within the animal, although the chemistry of butter fat and oil differ somewhat. Each is a suitable food. The vegetable oils, also, such as olive oil, palm oil, cocoanut oil, peanut oil, cottonseed oil, and others are all suitable foodstuffs. The digestibility of the various animal and vegetable fats is high." Some years later Carlson stated: "All the scientific data on the digestibility, flavor and color of the dietary fats show clearly that there is no significant difference in digestibility between animal and vegetable fats and that the acceptability of those fats in regard to color and flavor is a matter of past conditioning of the individual and of no other significance in nutrition."

*

Boutwell and others, after experimental studies on rats, concluded: "1. With lactose as the sole carbohydrate * * rats showed superior growth when fed butter or lard as compared to corn oil, cocoanut oil, cottonseed oil, soybean oil, peanut oil, olive oil and hydrogenated cottonseed oil. 2. With a mixture of carbohydrates composed of sucrose, starch, dextrose, dextrin and lactose in the diet, the average growth response of the animals fed vegetable oils was equal to that of the animals fed butter and lard. The growth rate on this ration was more rapid than when all the carbohydrate was present as lactose. 3. Properly fortified oleomargarine fats gave growth equal to butter fat over a period of 6 weeks when the above mixture or carbohydrates was incorporated in the rations." Deuel, on the other hand, found no difference in the growth of weanling rats at any time over a 12 week period whether they were fed mineralized skimmed milk powder, vitamin supplements and butter, or corn, cottonseed, peanut or soybean oils, or margarine. The extent of growth was confirmed at 3 and 6 weeks by roentgen determinations of length of the tibia. Also, the efficiencies of conversion of these various fats to body tissue were identical. These experiments refute the idea that butter fat possesses certain fatty acids not present in other fats, which are essential to growth.

5

The Council on Foods and Nutrition of the American Medical Association stated: "It is therefore possible to conclude that at present there is no scientific evidence to show that the use of fortified margarine in an average adult diet would lead to nutritional difficulties. A similar statement is probably justified in the case of growing children, but preliminary reports from animal experiments indicate that more work is necessary before any specific conclusions can be made." Graves' stated: "When pure, all fats are equally available for the energy needs of the body * * * the shortening powers and keeping qualities of [both butter and margarine] are about the same and they are equally assimilable." Again, Bloor pointed out: "Very little need be said about the relative nutritional value of fats and hence of availability and distribution for the reason that most of the ordinary food fats of both plant and animal origin consist mainly of the same few fatty acids―oleic, palmitic and stearic-in varying proportions, and it is to be expected that they would not differ much in digestibility or in metabolic usefulness."

8

Terms of

From the department of pediatrics, University of Illinois College of Medicine. Aided by a grant from the National Association of Margarine Manufacturers. the grant provided that findings from the study could be published, regardless of results. Holmes, A. D.: Digestibility of Oleomargarine, Boston M. and S. J, 192: 1210 (June 18), 1925.

2 Bunker, J. W. M.: Evidence Concerning the Reputed Health Values of Fats: A Review of the Literature. Am. J. Jub. Health 17: 997 (October) 1927.

3 Carlson, A. J. Facts and Fancies About Food Fats, Am. J. Pub. Health, 31: 1181 (November) 1941.

Boutwell, R. K.; Geyer, R. P.; Elvehjem, C. A., and Hart, E. B.: Further Studies on Comparative Value of Butter Fat, Vegetable Oils and Oleomargarine, J. Nutrition 26: 601 (December) 1943.

Deuel, H. J., Jr.: Movitt. E.; Hallman, L. F., and Mattson, F. Studies of the Comparative Nutritive Value of Fats: I. Growth Rate and Efficiency of Conversion of Various Diets to Tissue, J. Nutrition 27: 107 (January) 1944.

The Comparative Nutritional Value of Butter and Oleomargarine, report of Council on Foods and Nutrition, J. A. M. A. 119: 1425 August 22, 1942.

Graves, L. Fats in Our Daily Fare, Mod. Hosp. 60: 90 March 1943.

8 Bloor, W. R.: Role of Fat in the Diet, J. A. M. A. 119: 1018, July 25, 1942.

Recently Deuel' reaffirmed the fact that vitamin-fortified margarine has a nutritional value substantially equivalent to that of butter. This belief is supported by the conclusions of an entirely unprejudiced group, the committee on public health relations of the New York Academy of Medicine, which recommended in its report of February 1, 1943, that wide publicity, both lay and professional, be given to the fact that margarine fortified with vitamin A is nutritionally equal to butter. A similar conclusion was reached by the Food and Nutrition Board of the National Research Council in its Reprint and Circular Series No. 118, released in August 1943.

* *

Cowgill 10 concluded: "Edible fats, the melting points of which are not too high to prevent liquefaction in the alimentary tract, are digested and absorbed to about the same degree. Such differences as have been found are of no practical nutritional significance. * Natural fats differ with respect to their content of the essential unsaturated fatty acids but the amounts needed by the organism are so small that these are probably of no practical nutritional significance. Natural fats have not been found to differ appreciably in their effect on the body's needs for other dietary essentials. In a diet otherwise nutritionally satisfactory, a vegetable fat such as that contained in a margarine can serve adequately in place of butterfat for growth and reproduction, as shown by experiments with eight and more successive generations of rats."

11

*

*

*

13

Boer and colleagues recently reported on the presence of vaccenic acid in sumer butter. They stated that vaccenic acid has growth-promoting properties in rats. Euler and associates," however, reported inability to demonstrate any growth-promoting factor peculiar to butter. Moreover the same authors in later studies "to characterize the physiological influence on the rat's organism of different fats * * * compared the growth, the fertility, and the longevity of rats, which were reared on a fat nutrition, consisting either of butter or of margarine (fat mixture MW)." They concluded "that the nutritional and physiological value of the slightly varying fat mixtures, which we employed under the designation of MW and which correspond to the margarines manufactured in Sweden during the period from October 1943 to November 1946, is in no way inferior to the value of butter."

Deuel and associates,1 meanwhile found vegetable fats and margarine equal to or better than butter in supporting augmented growth in two series of tests with rats, based on administration of growth hormone from the anterior lobe of the pituitary body.

PLAN OF STUDY

Two groups of children were included in the study, which covered a period of 2 years. One group received only margarine as the table fat in the diet. It was used on bread and vegetables, as well as in the making of pastry and in frying. Fortified margarine was supplied by a number of the various companies now manufacturing the product and was the same as that sold to the public. The margarine used was all derived from vegetable fats, and contained no fat from animal sources. The second group of children used only butter for the same purposes. The margarine group lived in an institution housing 130 children ranging in age from 3 to 16 years. The children were half-orphans, for the most part, from broken homes. They attended the neighborhood schools and returned to the institution for lunch. Samplings also indicated that the children were served margarine on their occasional visits to families outside the institution.

The butter group was in another institution some 10 miles (16 kilometers) away. It included 125 children ranging in age from 6 to 17 years, who were mostly orphans.

Leuel, H. J., Jr.: The Butter-Margarine Controversy, Science 103: 183, February 15, 1946. 10 Cowgill, G. R. Relative Nutritive Values of Animal and Vegetable Fats, Physiol. Rev. 25 664, October 1945.

11 Boer, J. Jansen, B. C. P., and Kentie, A.: On the Growth-Promoting Factor for Rats Present in Summer Butter, J. Nutrition 33: 339, March 1947. Boer, J.; Jansen, B. C. P.; Kentie, A., and Knol, H. W.: The Growth-Promoting Action of Vaccenic Acid, ibid. 33: 359, March 1947.

12 von Euler, B.; von Euler, H., and Saberg, I.: Zur Kenntnis des Nahrwertes verschiedener Fette, Die Ernährung 7: 65, 1942; Versuche über die Nahrungsfaktoren der Butter, ibid. 8: 257, 1943.

13 von Euler, B., and von Euler, H.: Biological Facts on Vaccenic Acid, Ark. f. kemi, mimero. o. geol., June 1947, vol. 25B, no. 2.

14 Deuel, H. J., Jr.; Hendrick, C., and Crockett, M. E.: Studies on the Comparative Nutritive Values of Fats: VII, Growth Rate with Restricted Calories and on Injection of the Growth Hormone, J. Nutrition 31: 737, June 1946.

The diet in each institution was carefully supervised by trained dietitians and so regulated that 25 to 30 percent of the total calories were supplied by fat. The margarine constituted approximately 65 to 70 percent of the total fat calories. All of the children in both groups were weighed and measured each month under medical supervision. A careful check was made from time to time to be sure that weights and measurements were accurately determined. Routine redblood-cell counts and hemoglobin determinations were made on each child after the study was started and again 1 year later. Specialists in pediatrics supervised the medical care of all the children. A record was kept of the character and duration of all illnesses contracted by the children.

RESULTS

At the close of the study more than 200 records were available for analysis in the margarine group. Because of the turn-over in the institution, the period of study was not uniform and ranged from about a month to a maximum of 24 months. In the butter group there were about 150 records available, ranging up to 24 months' observation time on weight, height, red blood count, hemoglobin, and illnesses.

It was found that those children who stayed a very short time had extreme irregularities in weight gains. Since it would serve no useful purpose, and only increase the margin of error in the computations, all those records of children who were studied less than 6 months were discarded. About 40 in each group were so eliminated, leaving 160 records in the margarine group and 107 in the butter group.

Blood studies.-There were 65 children in the margarine group and 85 in the butter group from whose records differences in erythrocyte count and hemoglobin could be computed. At the beginning of the study there was an average red-cell count of 4,284,000 in the margarine group, the lowest being 3,710,000 and the highest 4,870,000. One year later the average was 4,198,000, the low 3,440,000 and the high 5,560,000. Comparable figures for the butter group were as follows: At the beginning the average red-blood-cell count was 4,629,000; low, 3,940,000; high, 5,640,000; 1 year later the average was 4,640,000; low, 3,920,000; and high, 5,560,000.

15

In the margarine group the average hemoglobin was 13.7 grams per hundred cubic centimeters, with a low of 11.3 and a high of 16.6. One year later the figures were 13.4, 10.5, and 18.6, respectively. For the butter group, at the beginning of the study the hemoglobin figures were: Average, 15.4 grams; low, 13 8; and high, 17.9. One year later the findings were: Average, 15.3; low, 12.4; and high, 17.5. All these figures are within normal limits in our clinical experience. Moreover, according to Osgood and Baker, there is an average of 12 grams of hemoglobin, ranging from 10 to 14 grams in the blood of normal children from 4 to 13 years of age. Griffith and Mitchell 16 stated that the red-blood-cell count remains at about 4,500,000 to 5,000,000 throughout childhood. Hence, it would appear that the blood-cell counts in both institutions were in line with counts found in similar groups of children. Furthermore, any variations which occurred were well within the usual errors in making hemoglobin and red-blood-cell determinations.

Height and weight.-All the records contained data on weight in pounds and ounces and height in inches. Gains in weight and height are partly a function of time, so a further adjustment had to be made to facilitate valid comparison. The basis of exactly 1 year was used for standardization. The period of the greatest majority of the cases was over a year. Since this means that the majority had to be reduced in value to be standardized, chance errors tend to be reduced as well.

An analysis of the total cases in each group is shown in tables 1 and 2. Average annual weight and height gains for boys and girls are shown. For additional comparison, the average gains for standard height-weight tables are included under the column headed "Standard."

Since changes in weight and height are also a function of age, only groups of like age are compared.

The standard error is appended to each value in the margarine group, and comparable value in the butter group. For a test of statistical significance at the

15 Osgood, E. E., and Baker, R. L.: Erythrocyte, Hemoglobin, Cell Volume and Color, Volume and Saturation Index Standards for Normal Children of School Age, Am. J. Dis. Child. 50: 343, August 1935.

16 Griffith, J. P C., and Mitchell, A. G.: The Diseases of Infants and Children, ed. 2, Philadelphia, W. B. Saunders Co., 1937, p. 38.

5-percent level, the difference between two comparable entries in the margarine and standard columns must be greater than twice the standard error. For instance, on the first line in table 1, under "Boys," in the column headed "Margarine," the average yearly gain in weight was 6.7 pounds (3,039 grams) with a standard error of 0.6 pound (272.2 grams). Twice 0.6 is 1.2, which, added to 6.7, gives 7.9 pounds (3,583.4 grams). The average gain in the standard column is 7.2 pounds (3,265.9 grams), well within the range of the standard error, and the difference is, therefore, not statistically significant.

TABLE 1.-Average yearly gain in weight (in pounds) for children observed at least 6 months

TABLE 2.-Average yearly gain in height (in inches) for children observed at

For a test of significance between a margarine and butter value the formula

is used. In only 3 cases does it seem necessary to perform this test. This is in table 1, under "Boys," ages 6-13 and ages 6-9, and table 3, under "Boys," 6-9 years of age. In none of these cases was the difference found significant.

In both table 1 and table 2 all average gains in the margarine group were at least as great (including the standard error) as the gain in either the butter group or a standard group, with two exceptions. These exceptions were in the 2- to 5-year age groups. This is due to the fact that the average age in this particular group in the margarine study was greater than the midvalue of the group. In fact, there were only one boy and one girl 2 years old included in the study. It is believed that the adjustment of all the values to an annual basis is accurate. However, to eliminate doubt as to the validity of the method, comparisons were made using only those subjects who were observed for the maximum time. There were 51 children in the margarine group who were studied the maximum time of 24 months. In the butter group there were 53 under observation for the maximum 24 months.

Again the values were reduced to an annual average gain, going far to reduce chance errors, since all values were practically halved. Tables 3 and 4 exhibit the data.

In only one case, the 2-5 year age group in the boys, in table 4, is the comparison unfavorable. In all other combinations margarine shows up as at least as efficient a source of fat for growth as butter in the normal diet.

Average yearly gain in height (in inches) in children observe

ore, it was noted that in the margarine group there was no inc unt of illness. Illnesses in general had been on the decline in group for the last 4 or 5 years, and this decrease in the inciden inued during the period of the study. This compares with condi he community for the last 5 years.

teresting to observe how the children accepted margarine. Wh its appearance on the table in its white form-and the study o the children, the younger children promptly accepted it. The older age group did not take to it too kindly. Very shortly therea shipment of colored margarine came in. This was cut up into and all the children then ate it readily and liberally. Thereaft served cut in pats and colored.

e during the period of the study was it considered that vitam definite role in the results of the study. The reason, of course, is garine used contained 15,000 units of vitamin A per pound-whi greater than the amount present in average butter.

dies showed that there were no significant differences between or butter groups.

ren in the margarine group experienced a high degree of good h study, and in comparing their health to that of the butter gro have been much better.

rmary records are compared it is readily seen that the marga much better than the butter group. We are not making any cl argarine group were healthier simply because their diet conta Other variables are more likely to account for their better he e during the course of the study did either institution experience lemic, and no doubt this may in part be due to the fact that practiced accepted preventive measures on all their children.