(Journal of Nutrition. 2001;131:409S-420S.)
© 2001 The American Society for Nutritional Sciences
Supplement
Infant Feeding in the 20th Century: Formula and Beikost1
Samuel J. Fomon
Department of Pediatrics, College of Medicine, University of Iowa, Iowa City, IA 52242
 |
ABSTRACT
|
|---|
The early years of the 20th century were notable for improvements in
general sanitation, dairying practices and milk handling. Most infants
were breast-fed, often with some formula feeding as well.
Availability of the home icebox permitted safe storage of milk and
infant formula, and by the 1920s, feeding of orange juice and cod liver
oil greatly decreased the incidence of scurvy and rickets. Use of
evaporated milk for formula preparation decreased bacterial
contamination and curd tension of infant formulas. From 1930 through
the 1960s, breast-feeding declined and cow’s milk and beikost were
introduced into the diet at earlier and earlier ages. Although
commercially prepared formulas, including iron-fortified formulas
replaced home-prepared formulas, few infants were breast-fed or
formula fed after 4–6 mo of age. Iron deficiency was prevalent. From
1970 through 1999, a resurgence of breast-feeding was associated
with a prolongation of formula feeding and an increase in usage of
iron-fortified formulas. By the end of the century, formula feeding
of older infants had largely replaced feeding of fresh cow’s milk and
the prevalence of iron deficiency had greatly decreased.
KEY WORDS: • infant feeding • breast-feeding • infant formula • beikost
|
INTRODUCTION
|
|---|
In 1948 when I began my pediatric residency, I had little awareness of
the history of infant feeding during the first half of the 20th
century. Since then, I have collected small bits of information in
relation to specific interests at the time. As a research fellow in
renal physiology, I studied osmotic diuresis and reviewed the
literature on renal solute load, including the reports concerning
infants. In the 1950s, when I was involved with Charlie May in
metabolic balance studies with normal infants, I reviewed the older
reports on intakes and excretion of protein, fat and some of the major
minerals. However, despite all the effort I invested over the years in
trying to define desirable characteristics of infant formulas, I did
not attempt an interpretive review of the history of infant feeding;
even in my 1993 book (Fomon 1993
), I relied primarily on
reviews by other authors for events that occurred before about 1950.
Thus, the thoughts presented here as they relate to the first half of
the 20th century require some indulgence of the readers. I am on firmer
ground during the second half of the century.
Among the most important advances in infant nutrition and feeding
during the latter half of the 20th century were the increase in
breast-feeding and the nutrition and feeding of preterm
infants—areas to be covered in other presentations in this symposium.
Although important advances were also made in feeding of infants with
chronic disease, my presentation is restricted to nutrition and feeding
of normal term infants.
A number of changes in public health during the latter half of the 19th
century contributed to more successful formula feeding of infants in
the early part of the 20th century (Fomon 1993
). Most
important of these were improvements in general sanitation, disposal of
garbage and, at least in some cities, chlorination of water. Handling
and storage of milk improved. The biochemical differences between the
major components of human milk and cow’s milk had been defined. Even
at the beginning of the 20th century, it was generally appreciated that
infant formulas based on cow’s milk required the addition of water and
carbohydrate. A number of commercially prepared formulas were patented.
Liebig’s food for infants, marketed in 1867 as a liquid and
subsequently as a powder, consisted of wheat flour, cow’s milk, malt
flour and potassium bicarbonate (Forsyth 1910–1911
, Smith 1885
). Other formulas were introduced in rapid succession;
by 1883, 27 brands of patented infant foods were available
(Bracken 1953
). However, relatively few infants were fed
commercially prepared formulas.
|
The years 1900–1930
|
|---|
In 1900, general sanitation and dairying practices, although
better than in the past, were still quite primitive by current
standards. Seamless rubber nipples, which could be fitted over the
necks of feeding bottles had become available but were apparently not
widely used (Brennemann 1912
); thus, cleaning of bottles
and nipples was generally unsatisfactory. Means for safe storage of
formula in the home were not generally available, and adulteration of
milk was common. The possibility that inadequate intakes of vitamins or
of trace minerals could lead to disease had not been considered. As
stated by Langworthy in 1898 (McCollum 1957
), it was
generally believed that "Foods have a dual purpose: Building and
repair. Energy for heat and work. Foods consist of the nutrients
protein fat and carbohydrates and various mineral salts."
Although reliable data on the prevalence of breast-feeding during
the early years of the 20th century are not available, comments in the
literature suggest that most infants were breast-fed throughout
most of y 1 of life, many of them also being fed some formula [see,
e.g., Friedenwald and Ruhrah (1905)
]. A survey of a
number of urban centers in 1912–1919 indicated that at 12 mo of age,
13% of infants were exclusively breast-fed and 45% were partially
breast-fed (Yankauer 1994
).
There is reason to believe that formula feeding in the early 1900s was
less successful in the United States than in Europe. In Europe, at
least in Germany, milk was almost uniformly boiled for use in infant
formulas, whereas in the United States, raw milk was most commonly used
(Brennemann 1911
). The strong prejudice against the use
of heat-treated milk was based on the observation that scurvy
occurred primarily in infants fed sterilized, condensed or pasteurized
milk (American Pediatric Society 1898
).
The early years of the 20th century were notable in the United States
for the adoption by many physicians of a complex sequence of changes in
formula composition, the "percentage method" or "American
method" of formula feeding (Rotch 1907
). The aim was
to provide a formula with a composition close to that of human milk but
taking into account the digestive capability of the individual infant.
Nevertheless, the major emphasis was on the ratios of protein, fat and
carbohydrate and not on energy density; formulas commonly ranged from
<50 kcal/dL to >80 kcal/dL. Formula preparation was so complex that
it was commonly performed in commercial laboratories dedicated to this
purpose. With the endorsement of Rotch, the Walker-Gordon Farm had
been established in 1891 for the production of clean milk (Morse 1935
) and Walker-Gordon Laboratories, which used this milk,
were in operation in many cities in the early 1900s (Friedenwald and Ruhrah 1905
, Morse 1935
). In retrospect,
although the system of formula preparation was unnecessarily complex,
the formulas prepared by the Walker-Gordon Laboratories were made
with care, were unlikely to be seriously contaminated with pathogens
and were therefore generally more satisfactory than formulas made in
the home. For all its prominence in the literature, this "American
Method" of formula prescription was not widely used in rural areas
nor by less affluent families in urban areas. Thus, the majority of
formula-fed infants received formulas made in the home from whole
milk or "top milk" (i.e., milk with 7–10% fat). Because the tough
curd formation associated with feeding cow’s milk protein presented a
greater problem to the infant than the digestion of butterfat, the use
of "top milk" resulted in more digestible formulas.
By 1912, clean milk was generally available in New York City
(Rosen 1958
) and it seems likely that similar
improvements in dairying practices and milk handling had occurred in
much of the United States. Rubber nipples that could be readily cleaned
had come into widespread use (Brennemann 1912
), and safe
storage of milk in many homes had become possible because of the
availability of the kitchen icebox (Fig. 1
). The important contribution on the energy requirements of infants
(Rubner and Heubner 1899
) had gained general recognition
and at least some physicians recommended an energy intake of 100
kcal/(kg · d) during the first few months of life and somewhat
lesser intakes per unit of body weight subsequently (Brennemann 1912
). Nevertheless, even in the 1920s, formulas varied
considerably in energy density. Those fed from 1924 to 1929 to infants
under the care of the Infant Welfare Society of Chicago (Grulee et al. 1934
) were no >53 kcal/dL, whereas formulas made from
whole milk with added Karo syrup (Marriott and Davidson 1923
) or from evaporated milk diluted 1:1 with water and the
addition of Karo syrup (Marriott 1927
) provided nearly
100 kcal/dL.
View larger version (10K):
[in this window]
[in a new window]
|
Figure 1. Changes in infant feeding and nutrition from 1900 to 1930: major
improvements in sanitation, dairying practices and milk handling
occurred from 1900 to 1912; sterilizable rubber nipples and the home
icebox became widely available by 1912; beginning in the early 1920s,
feeding of orange juice and cod liver oil resulted in decrease in the
prevalence of scurvy and rickets, and use of boiled or evaporated milk
in formula preparation from the mid-1920s resulted in infant formulas
with lower curd tension and less bacterial contamination.
|
|
|
Vitamins
|
|---|
As reviewed by McCollum (1957)
, a number of animal
studies in the late 1800s and early 1900s had demonstrated that,
despite the scientific consensus of the time, diets containing protein,
fat, carbohydrate and mineral salts were inadequate to support life. An
enlightened view was presented by Hopkins (1906)
: "The
animal body is adjusted to live either on plant tissues or on other
animals, and these contain countless substances other than proteins,
carbohydrates and fats. Physiologic evolution, I believe, has made some
of these well nigh as essential as are the basal constituents of the
diet." In 1912, Funk (1912)
suggested that beriberi,
scurvy, pellagra and possibly rickets were caused by deficiency in the
diet of special substances for which he proposed the name,
"vitamines."
Scurvy.
Although scurvy of the adult had been recognized as early as 1734
[citation of Bachstrom by Stewart and Guthrie (1953)
],
the relation between infantile scurvy and scurvy of the adult was slow
to be recognized. It was largely through the efforts of Hess that it
became customary in the 1920s to supplement the diets of infants with
fruit or vegetable juices (McCollum 1957
). The
prevalence of infantile scurvy then decreased substantially.
Rickets.
As urban living in the United States increased during the late 1800s
and early 1900s, infantile rickets increased. The Russian pediatrician,
Schabad, in a series of reports published between 1908 and 1912,
demonstrated that cod liver oil was effective in curing and preventing
rickets (Holt 1963
). In 1920, Mellanby (1920)
demonstrated that a fat-soluble substance could
prevent rickets in puppies; in 1922, McCollum and co-workers (1957)
demonstrated that the fat-soluble substance was not
vitamin A. Use of cod liver oil as a prophylactic measure against
rickets became widespread in the United States by the mid-1920s.
|
Other developments before 1930
|
|---|
By the mid-1920s, when it was known that infantile scurvy could be
prevented by daily feeding of fruit juices, the prejudice against use
of boiled milk in infant formulas disappeared and formula feeding
became much more successful. To modify curd tension, lactic acid was
commonly used. The use of lactic acid rather than lime water for
modifying curd tension may have occurred because of the advantage of
acidified over alkalinized formulas in inhibiting bacterial growth.
Evaporated milk was first marketed by Gail Borden in 1858
(Wharton 1941
); beginning in 1885, it was sold in
hermetically sealed cans sterilized by heat. However, because of fear
of producing scurvy, it was not used in infant feeding until the 1920s,
when its use was promoted by several of the leading pediatricians of
the time (Brenneman 1929
, Marriott 1927
,
Marriott and Schoenthal 1929
). Evaporated milk was
relatively inexpensive, could be stored at room temperature and was
free of bacterial contamination until the can was opened. The processes
of evaporation, homogenization and heat treatment resulted in physical
changes in the milk, with an increased percentage of casein adsorbed to
the surface of the fat globules (Council on Foods 1937a
), thus contributing to the reduction in curd tension.
|
Beikost
|
|---|
Although cereal was commonly included as a constituent of infant
formulas in the early 1900s, the purpose of its inclusion was the
reduction of curd tension, not as an energy source. With the use of
evaporated milk for infant formulas, cereal was no longer needed. On
the basis of her review of the literature, Adams (1959)
stated that until the 1920s, the usual feeding pattern in the United
States was introduction of sieved vegetable soup by the end of y 1,
potato at 
18 mo and other vegetables not until 2 y of age or
later. She pointed out that in Holt’s The Diseases of Infancy
and Childhood, the recommended age for introduction of green
vegetables in the 1911 edition was 36 mo, but that by the 1929 edition,
the age had decreased to 9 mo. However, it is evident that earlier
introduction of beikost was common at least in some areas. Infants
under the care of the Infant Welfare Society of Chicago between 1924
and 1929 received cereal at 5 mo of age and a vegetable at 6 mo of age
(Grulee et al. 1934
).
|
The years 1930–1970
|
|---|
Various developments in infant feeding and nutrition from
1930–1970 are indicated in Figure 2
. Although data on the percentage of infants who were breast-fed in
the United States from 1930 to 1950 are less satisfactory than later
data, there is no question that the trend was downward. Data from the
National Fertility Study (Hirschman and Hendershot 1979
,
Hirschman and Butler 1981
), indicate that from 1931 to
1935, >70% of first-born infants and a somewhat lesser percentage
of second-born infants were initially breast-fed and 40% of
infants were breast-fed for at least 6 mo. By 1946–1950, initial
breast-feeding of first-born infants had decreased to 50% and
only 20% were breast-fed for at least 6 mo. A survey of hospitals
carried out in 1945 (Bain 1948
) indicated that 69% of
infants discharged 
7 d after birth were breast-fed and 60% of
infants discharged >7 d after birth were breast-fed. The data of
Bain regarding percentage of infants initially breast-fed are based
on a review of discharge records and therefore are likely to be more
accurate than the recall data of Hirschman and co-workers
(Hirschman and Hendershot 1979
, Hirschman and Butler 1981
). During the 1950s and 1960s, the trend in
breast-feeding was steadily downward, and by the early 1970s, only
25% of infants were breast-fed at age 1 wk and only 14%
between 2 and 3 mo of age (Fig. 3
).
View larger version (13K):
[in this window]
[in a new window]
|
Figure 2. Changes in infant feeding and nutrition from 1930 to 1970: breast
feeding declined, whereas beikost and fresh cow’s milk were introduced
at earlier and earlier ages. The first federal regulations concerning
infant formulas went into effect in 1941 and, beginning in 1950,
commercially prepared formulas began to replace home-prepared
formulas. Iron-fortified formulas were introduced in 1959.
|
|
View larger version (12K):
[in this window]
[in a new window]
|
Figure 3. Exclusively and partially breast-fed infants as percentage of all
infants in 1971, 1991 and 1998. Data for 1971 from Martinez and Krieger (1985)
and for 1991 from personal communication from
Ross Mothers’ Survey (Greenbaum, S., 1992, Ross Products Division,
Columbus, OH). A smoothed curve is presented for 1998 based on data
from personal communications from Boettcher, J. A. (1999, Mead
Johnson Nutritionals, Evansville, IN) and the Ross Mothers’ Survey
(Ryan, A., 2000; Ross Products Division, Columbus, OH).
|
|
|
Home-prepared infant formulas
|
|---|
From the 1930s or early 1940s, most formulas fed to infants in the
United States were prepared by mixing evaporated milk or fresh cow’s
milk with water and adding carbohydrate. A typical evaporated milk
formula, as prepared in 1949, when I was a pediatric resident, included
1 can (13 fl oz) of evaporated milk, 19 fl oz of water, and 1 oz of
carbohydrate, usually in the form of corn syrup (Karo) or sucrose. Such
a formula provided 67 kcal/dL, with 15% of energy from protein,
42% from carbohydrate and 43% from fat. Home-prepared formulas
were sometimes made with cow’s milk (usually pasteurized and
homogenized) rather than with evaporated milk. These formulas provided
about the same distribution of energy from protein, fat and
carbohydrate as did the evaporated milk formulas. Most evaporated milk
and most pasteurized, homogenized whole cow’s milk were fortified with
vitamin D. Orange juice was given as a source of vitamin C.
Improved general sanitation, safe supplies of water and milk, and
better understanding of both microbiology and nutrient requirements
resulted in a high degree of success with formula feeding, and it was
the opinion of most physicians and the general public that formula
feeding was about as safe and satisfactory as breast-feeding.
However, the infant formulas in general use in the 1950s were
associated with a number of problems unappreciated by physicians and
parents, including the following: 1) the high potential
renal solute load placed the infants, especially young infants, at risk
of developing hypernatremic dehydration during illness (Fomon and Ziegler 1999
); 2) the low content of iron in the
formulas together with the high intake of inhibitors of iron absorption
(Fomon 1993
) were responsible for a high prevalence of
iron deficiency and, in the case of whole-milk formulas, probably
with the added problem in some infants of increased intestinal blood
loss (Ziegler et al. 1990
); 3) intakes of
essential fatty acids were low. In addition, scurvy continued to be
seen. A survey of 226 teaching hospitals in the United States indicated
that during the years 1956–1960, 713 infants and children were
admitted to these hospitals because of scurvy (Committee on Nutrition 1962
).
|
Commercially prepared formulas
|
|---|
Milk-based formulas.
From the late 1800s, a number of commercially prepared formulas were
available in the form of powders that merely required the addition of
water before being ready to feed to infants. Many of these formulas had
been developed in an attempt to mimic the chemical composition of human
milk, and several researchers had focused their attention on the
greater percentages of low-molecular-weight fatty acids in cow’s milk
than in human milk, believing that these were responsible for the poor
tolerance of infants to butterfat (Gerstenberger et al. 1915
). Thus, even in the early 1900s, formulas free of
butterfat had been marketed. The cost of powdered formulas was
appreciably greater than that of formulas made from evaporated milk or
whole cow’s milk, and usage of commercially prepared formulas was
rather low. However, beginning in 1951, when concentrated liquid
formulas (133 kcal/dL) were introduced, considerations of convenience
began to supersede considerations of cost, and the popularity of
commercially prepared formulas increased dramatically
(Fig. 4
). By 1960, concentrated liquid formulas had largely replaced powdered
formulas (Fig. 4)
. The change from home-prepared formulas to
commercially prepared formulas was accelerated by the introduction in
1959 of iron-fortified formulas and the vigorous promotion of these
formulas by the formula industry and by pediatricians (Andelman and Sered 1966
, Committee on Nutrition 1971
). By
the late 1960s, <10% of infants were fed home-prepared formulas
(Fig. 5
).
View larger version (15K):
[in this window]
[in a new window]
|
Figure 4. Commercially prepared formula sales by type from 1950 to 1998: powder,
concentrated liquid (133 kcal/dL) and ready-to-feed (67 kcal/dL). Data
through 1987 based on Ross Mothers’ Survey, personal communication
from Martinez, G. A. (1989, Ross Products Division, Columbus, OH);
subsequent data based on personal communication from Boettcher, J. A. (2000, Mead Johnson Nutritionals, Evansville, IN).
|
|
View larger version (14K):
[in this window]
[in a new window]
|
Figure 5. Percentage of 2- to 3-mo-old infants receiving various forms of feeding
from 1958 to 1970: feeding of commercially prepared formulas (prepared)
increased and feeding of formulas based on evaporated milk (EM)
decreased. Relatively few infants were breast-fed and few young
infants were fed cow’s milk (CM). Data based on Ross Mothers’ Survey,
personal communication from Martinez, G. A (1989, Ross Products
Division, Columbus, OH).
|
|
From at least the 1930s (Powers 1935
) until the 1950s,
the protein concentration of human milk was believed to be greater than
is now known to be the case, and many pediatricians believed that
cow’s milk protein was so inferior to human milk protein for meeting
the needs of infants that infants fed formulas required a considerably
greater intake of protein than did breast-fed infants. Protein
content of a number of widely used formulas ranged from 3.3 to 4.0
g/100 kcal and several formulas recommended for use in management of
infants with diarrhea provided 5.7–6.3 g/100 kcal (Fomon 1967
). During the late 1950s and early 1960s, most of the
leading commercially prepared formulas fell into one of two classes.
One class (e.g., Lactum, Mead Johnson) consisted of formulas similar to
home-prepared evaporated milk formulas but with added vitamins; the
other class (e.g., Similac and SMA) were of lower protein content and
contained a mixture of vegetable and oleo oils with added vitamins and
minerals. The gradual takeover of the market by the latter formulas
seems not to have been based on considerations of nutrient requirements
or renal solute load, but on the unpleasant odor of regurgitated
butterfat after its partial digestion and on the impression that
formulas similar to the home-prepared evaporated milk formulas led
to constipation.
As early as 1923, James Gamble had gained at least some understanding
of renal excretion of solutes (Abt 1965
), but it was not
until the 1950s that the relation of renal solute load to water balance
in infants received serious consideration (Cooke et al. 1950
, Darrow et al. 1954
, Pratt et al. 1948
), and not until the 1960s that renal solute load began to
be considered in the design of infant formulas. However, even at the
end of the century, infant formula regulations permitted the marketing
of formulas with undesirably high potential renal solute load. In 1998,
an Expert Panel recommended to the Food and Drug Administration (FDA)
that infant formulas provide a potential renal solute load (i.e.,
solutes of dietary origin that would require renal excretion if none
were diverted into growth and none were lost through nonrenal routes)
no >33 mosm/100 kcal (Raiten et al. 1998
). Although the
FDA took no immediate action on this recommendation, the upper limit of
33 mosm/100 kcal is well above that of formulas marketed during the
last 20 y of the 20th century. However, most of the formulas fed
in the first half of the 20th century exceeded this maximum and
undoubtedly contributed to the prevalence of hypernatremic dehydration.
An iron-fortified formula was introduced in the United States in
1959; by the mid 1960s, most manufacturers offered the same base
formula with or without substantial iron fortification (Fomon 1967
). Many parents and physicians were reluctant to use
iron-fortified formulas because they believed that feeding such
formulas was responsible for constipation, fussiness and intestinal
disturbances in the infants. Studies that failed to confirm such
adverse effects (Nelson et al. 1988
, Oski 1980
) did not seem to change the prejudice against
iron-fortified formulas.
Recognizing that human milk has a preponderance of whey proteins,
whereas cow’s milk has a preponderance of caseins, a milk-based
formula with a whey/casein ratio similar to that of human milk was
introduced in the United States in 1962 and by the mid 1990s,
whey-predominant formulas became the rule. Nevertheless, the whey
proteins of cow’s milk are quite different from those of human milk;
even today, rather meager evidence exists that a milk-based formula
with added whey proteins results in a product that is superior to a
milk-based formula without the additional whey proteins.
Nonmilk-based formulas.
A formula based on soy flour was developed by Hill as a feeding for
infants allergic to cow’s milk and became commercially available in
1929 (Abt 1965
). Formulas prepared from soy flour were
pale tan in color and had a nutty odor. Parents complained that the
formulas produced loose, somewhat malodorous stools, and resulted in
staining of the reusable cloth diapers that were in general use.
Excoriation of the diaper area was common. The stool characteristics
resulted primarily from the presence of considerable amounts of fiber
in the soy flour. In addition to soy-based formulas, a
meat-based formula and a casein hydrolysate formula were marketed.
Several of these special formulas were not fortified with vitamins when
initially marketed, apparently because pediatric allergists believed
that the vitamin mixes used for vitamin fortification of formulas might
include allergens. In the 1950s and 1960s, a number of vitamin
deficiencies were described (Fomon 1993
). In addition,
goitrogens present in soy flour were responsible for development of
goiters in infants fed a soy flour–based formula unfortified with
iodine. A few cases of vitamin K deficiency were reported in the 1940s
in infants fed a meat-base infant formula (protein from beef heart)
or a casein hydrolysate formula before these formulas were fortified
with vitamin K (Fomon 1993
).
Formulas prepared with isolated soy protein became commercially
available in the United States in the mid-1960s and within 10 y
almost completely replaced soy flour–based formulas. Isolated soy
protein–based formulas are similar in color to milk-based formulas
and are nearly odorless. Because most of the fiber is removed during
the protein isolation process, the infant’s stools are generally
similar to those of infants fed milk-based formulas. However, the
process employed in isolation of the protein resulted in elimination of
most of the vitamin K that had been naturally present in the soy
flour–based products, and a few cases of vitamin K deficiency were
reported before the products were fortified with vitamin K
(Fomon 1993
). Development of nutrient deficiencies in
infants fed milk-free formulas was responsible in part for the
development of a series of federal regulatory actions on nutrient
content of infant formulas.
|
Governmental regulations concerning infant formulas
|
|---|
The United States was among the last of the major industrialized
countries to implement federal regulations concerning safety of infant
formulas (Miller 1989
). It was not until 1938 that the
new Food and Drug Act included reference to foods for special dietary
purposes, including infant formulas. In 1941, the FDA declared that a
food sold for use by infants should include a label declaration for
moisture, energy, protein, fat, available carbohydrates, fiber,
calcium, phosphorus, iron and vitamins A, B-1, C and D.
In 1952 and 1953, an alteration in the method of heat treatment of
concentrated liquid SMA resulted in a decrease in vitamin B-6 content,
and clinical manifestations of vitamin B-6 deficiency developed in a
number of infants (Fomon 1993
). As a result of this
experience, the FDA in 1962 published a proposed revision of the 1941
regulations. A revised final regulation published in 1966 included the
requirement for minimal levels of 11 vitamins and minerals; because of
controversy over the regulations, however, it was not put into effect
(Miller 1989
). Instead, the FDA asked the Committee on
Nutrition of the American Academy of Pediatrics (AAP) to recommend
levels of nutrients in infant formulas. The report of the
Committee on Nutrition (1967)
was used as a basis for
public hearings in 1968–1969, and the final regulation, published in
1971 (FDA 1971
) included minimum requirements for
protein, fat, linoleic acid and 17 vitamins and minerals.
Commercial formula services and the development of ready-to-feed
formulas.
Throughout the first half of the 20th century, hospitals maintained
formula laboratories to prepare formulas for newborns and other
formula-fed infants. This activity required special equipment, was
labor intensive and presented formidable problems in quality control.
In the early 1950s, commercial formula services began operating in a
number of metropolitan areas in the United States (Committee on Nutrition 1965
) and many hospitals elected to use these
services rather than to continue their own activities in formula
preparation. By the early 1960s, considerable discussion centered about
the cost effectiveness of purchasing ready-to-use formulas from outside
sources rather than preparing them intramurally (Fomon 1993
). It was evident that use of a commercial formula service
influenced the choice of stock formula selected by a hospital. For
example, if an evaporated milk formula was offered at $0.09/bottle
while a commercially prepared formula was offered at $0.12/bottle (the approximate purchase price in the early 1960s), the less expensive
formula was likely to be chosen. Manufacturers of various prepared
formulas were therefore motivated to develop competing feeding systems.
In 1963, the Mead Johnson Company introduced the Beneflex system of
feeding in which bulk quantities of any infant formula manufactured by
that company could be transferred aseptically to feeders suitable to
the needs of individual infants (Fomon 1993
). Soon
afterward, the formula manufacturers were able to offer sterile
ready-to-feed formulas in disposable bottles with disposable or
reusable nipples. These were first used in hospitals but were
subsequently made available to the general public. An indication of the
rapid rise in sales of ready-to-feed formulas during the late 1960s and
early 1970s may be seen from Figure 5
. The data in the figure apply to consumer sales and do not include
hospital usage. Early in 1965, approximately equal numbers of hospitals
in the United States used ready-to-feed formulas supplied by
manufacturers and formulas supplied by locally operated commercial
formula services. By 1970, nearly all of the locally based commercial
formula services had ceased to exist, few hospitals prepared their own
formulas intramurally and most newborn nurseries used commercially
prepared, ready-to-feed formulas.
|
Cow’s milk
|
|---|
Comprehensive data on the percentage of infants fed cow’s milk at
various ages in the 1940s through the 1960s are not available, but
Harris and Chan (1969)
in a limited survey found that
60% of infants were fed whole milk by 4 mo of age. The percentage may
not have been quite as high for the entire country. In 1971, >30% of
infants from 3 to 4 mo of age, >40% of infants from 4 to 5 mo of age
and 60% of infants from 5 to 6 mo of age were fed cow’s milk
(Fig. 6
). On the basis of data from various sources, I estimated that in 1975
(Fomon 1975
), 58% of 5- to 6-mo-old infants were fed
cow’s milk.
View larger version (13K):
[in this window]
[in a new window]
|
Figure 6. Infants fed cow’s milk (no breast-feeding, no formula feeding) as
percentage of all infants in 1971, 1980 and 1991. Data for 1971 and
1980 from Martinez et al. (1981)
; data for 1991 based on
personal communication from Ross Mothers’ Survey (Greenbaum, S., 1992,
Ross Products Division, Columbus, OH).
|
|
Because it was not yet appreciated that feeding of homogenized,
pasteurized cow’s milk to young infants could predispose to
dehydration during illness and to development of iron deficiency, there
seemed therefore little reason not to change at an early age from
feeding formula to feeding fresh cow’s milk. Cow’s milk was
considerably less expensive than infant formula, required no mixing and
was a staple item in the home. Moreover, many parents probably
considered that the ability of an infant to tolerate at a young age a
diet more closely approaching that of older children was an index of
infant development and maturity.
|
Beikost
|
|---|
Following the trend of the previous two decades, the recommended
age for introduction of beikost continued to decrease from 1930 to the
early 1970s. In 1935, Marriott (1935)
suggested mo 5 or
6 as an appropriate age for introduction of solid foods, and in 1937
the American Medical Association (Council on Foods 1937b
) stated that pediatricians favored feeding of strained
fruits and vegetables at 4–6 mo of age. Beal (1957)
reported that in an upper socioeconomic group in Denver, strained foods
were offered to the infant at increasingly early ages during the years
1946 through 1955, and this seemed to be a general trend. Among
pediatricians responding to a survey in 1954, feeding of solids was
recommended before 8 wk of age by 66% and before 3 mo of age by 88%
(Butler and Wolman 1954
). Most extreme were the
recommendations of Sackett (1953)
, who promoted feeding
of cereal at 2–3 d of age, strained vegetables at 10 d and
strained fruits at 17 d. In 1963, Epps and Jolley (1963)
reported that when infants were seen for a first health
visit at 1–2 mo of age in the Child Health Clinics of the District of
Columbia, 83% were already receiving beikost. On the basis of a survey
in Rochester, Minnesota, in the late 1960s, Harris and Chan (1969)
reported that nearly 80% of infants were being fed
cereal by 1 mo of age. In 1975, I estimated that 5- to 6-mo-old infants
obtained 40% of energy intake from beikost (Fomon 1975
). Such a diet can be calculated to be generous in protein
and carbohydrate and relatively low in fat (Fomon et al. 1990
). Because few infants were fed iron-fortified formulas
(or any formulas) after 5 or 6 mo of age, beikost contributed most of
the dietary iron for most infants, and dry powdered cereals fortified
with iron were the major contributors.
In the 1940s and 1950s, infant cereals were fortified with sodium iron
pyrophosphate or other insoluble iron compounds of low bioavailability;
beginning in 1972, the cereals were fortified with electrolytic iron
powder (Committee on Nutrition 1976b
). On the basis of a
report by Rios et al. (1975)
that electrolytic iron
powder was as well absorbed by infants as was ferrous sulfate, it was
generally assumed that regular feeding of iron-fortified cereals
could meet the infant’s need for iron. Therefore, most physicians saw
no objection to feeding cow’s milk.
Throughout the 1960s, salt, monosodium glutamate, sugar and modified
food starches were included in the preparation of many commercially
available strained and junior foods. Salt, monosodium glutamate and
sugar were presumably added to satisfy the preferences of adult taste
panels, and the modified food starches were used to achieve and
maintain the desired physical appearance, consistency and texture of
the products. The manufacturers voluntarily discontinued the use of
monosodium glutamate in 1969. In 1970, a subcommittee of the Food
Protection Committee, Food and Nutrition Board, National Academy of
Sciences/NRC recommended an upper limit of 0.25% for salt added to
commercially prepared infant foods (Filer 1971a
) and
concluded that, when used in accordance with federal regulations, there
was no toxicologic basis for excluding modified food starches from the
diets of infants (Filer 1971b
). Over the next few years,
the manufacturers adjusted their formulations to decrease the
concentration of salt in infant foods. The downward trend in addition
of salt was accompanied by a downward trend in addition of sugar. By
1977, the addition of salt had been discontinued, and sugar was added
to fewer products and in smaller amounts than previously. The decrease
in addition of sugar resulted in a considerable decrease in the energy
density of some products, e.g., strained fruits provided an average of
82 kcal/100 g in 1972 and only 54 kcal/100 g in 1984 (Anderson and Ziegler 1987
). By the late 1970s, all manufacturers had
reduced the number of beikost items to which modified food starches
were added, and had discontinued use of all but a few types of modified
starches.
|
American Academy Committee on Nutrition
|
|---|
In 1954, soon after I arrived in Iowa as an assistant professor in
the Department of Pediatrics, the chairman of the department, Charles
D. May, was asked by the executive director of the AAP to serve as
chairman of a new committee of the Academy, the Committee on Nutrition,
which had been established by the Executive Board of the Academy on
April 1, 1954 (Executive Board 1956
). In those days, the
AAP operated in a much less formal mode than was to be the case in
later years, and May was given complete freedom in choosing members,
mostly from pediatric departments, to serve on the committee. A liaison
group of scientists and administrators was also established with
individuals from governmental agencies, including the FDA and the
infant food industry. With the aid of suggestions from May, the
Executive Board of the Academy in 1956 outlined the scope of the
Committee activities as follows: "This Committee shall concern itself
with standards for nutritional requirements, optimal practices and the
interpretation of current knowledge as these affect infants, children
and adolescents."
The first report of the Committee, "Ethics and etiquette in
advertising" (Committee on Nutrition 1956
) was written
by May. The second report, "Water requirement in relation to osmolar
load as it applies to infant feeding" (Committee on Nutrition 1957
), I prepared at May’s request. Fortunately, the report
did not acknowledge that I was the author because I had included
calcium and magnesium as components of the renal osmolar load. It was
the first of a series of gradually improving statements that I, and
later Ziegler and I, published on the topic. The second chairman of the
Committee was Charles U. Lowe (1957–1960) and I was the third chairman
(1960–1962). During those early years, the reports of the Committee
were primarily educational and did not include policy statements. It
was not until the mid 1960s that the Committee finally gained
nutritional prominence through its assistance to the FDA in defining
nutritional requirements for infant formulas and in setting policy for
nutritional practices relating to infants, children and adolescents.
During the last quarter of the 20th century, the Committee on Nutrition
exerted an enormous influence on child nutrition, most notably on
aspects of infant feeding.
|
The years 1970–1999
|
|---|
Infant feeding in the United States during the last 30 y of
the 20th century was marked by increases in breast-feeding and
formula feeding and a decrease in feeding of cow’s milk (Fig. 7
). The increase in breast-feeding in industrialized countries in the
1970s was worldwide, and the reasons for the increase after several
decades of decline are not easy to identify. The movement toward
increased breast-feeding seemed to arise from the general public
rather than from health professionals, and may have been in part
associated with negative publicity directed against the formula
industry. The formula industry was accused of interfering with
breast-feeding in lesser industrialized countries by its aggressive
marketing of infant formulas (Joseph 1981
,
McComas 1988
). In the 1970s in the United States, the
National Council of Churches’ Interfaith Center on Corporate
Responsibility and the Infant Formula Action Coalition mounted an
effective public awareness campaign. Likely in response to this new
climate, the infant formula manufacturers increased their efforts to
promote breast-feeding.
View larger version (14K):
[in this window]
[in a new window]
|
Figure 7. Changes in infant feeding and nutrition from 1970 to 2000:
breast-feeding increased, although the duration of
breast-feeding was generally <6 mo and was commonly accompanied by
formula feeding. The change from breast- or formula feeding to feeding
of cow’s milk occurred at progressively later ages. The women, infants
and children (WIC) program included a larger and larger percentage of
infants from low income families and contributed to increased
breast-feeding and use of iron-fortified formulas in this
group. The American Academy of Pediatrics (AAP) Committee on Nutrition
provided guidance to the Food and Drug Administration (FDA) on the
nutrient content of infant formulas and issued policy statements on
infant feeding, including one that contributed to the later
introduction of cow’s milk. Federal regulations on infant formulas
included the Infant Formula Act of 1980 and a new regulation in 1985 on
nutrient content of infant formulas.
|
|
|
Federal regulations regarding infant formulas
|
|---|
&
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
TOP
ABSTRACT
INTRODUCTION
