OBJECTIVE. Infancy is a time of rapid transition from a diet of virtually nothing but milk (either breast milk or infant formula) to a varied diet from nearly all food groups being consumed on a daily basis by most infants. Despite various recommendations about infant feeding, little is known about actual patterns of feeding among US infants. This article documents transitions in infant feeding patterns across the first year of life and determinants of key aspects of infant feeding.
METHODS. Using data from the Infant Feeding Practices Study II, we analyzed responses to a 7-day food-recall chart that was administered every month. The sample size declined from 2907 at birth to 1782 at 12 months of age.
RESULTS. Although 83% of survey respondents initiated breastfeeding, the percentage who breastfed declined rapidly to 50% at 6 months and to 24% at 12 months. Many of the women who breastfed also fed their infants formula; 52% reported that their infants received formula while in the hospital. At 4 months, 40% of the infants had consumed infant cereal, 17% had consumed fruit or vegetable products, and <1% had consumed meat. Compared with infants who were not fed solid foods at 4 months, those who were fed solid foods were more likely to have discontinued breastfeeding at 6 months (70% vs 34%) and to have been fed fatty or sugary foods at 12 months (75% vs 62%).
CONCLUSIONS. Supplementing breast milk with infant formula while infants were still in the hospital was very common. Despite recommendations that complementary foods not be introduced to infants aged 4 months or younger, almost half of the infants in this study had consumed solid foods by the age of 4 months. This early introduction of complementary foods was associated with unhealthful subsequent feeding behaviors. Pediatrics 2008;122:S36–S42
- AAP =
- American Academy of Pediatrics •
- IFPS =
- Infant Feeding Practices Study •
- WIC =
- Special Supplemental Nutrition Program for Women, Infants, and Children
The American Academy of Pediatrics (AAP) recommends that infants be breastfed exclusively for the first several months of life and that breastfeeding should continue through the first year of life. The AAP Committee on Nutrition recommends that infants begin consuming foods in addition to breast milk or formula after 4 months of age, preferably at 6 months of age.12 Early foods should include a source of iron, either fortified infant cereal or meat, but foods high in fat or sugar (eg, sugar-sweetened beverages, French fries, and candy) should not be given to infants.3 Within these constraints, infants should be introduced to a variety of foods as they transition to table foods.4
Little is known about current patterns of feeding among US infants during their first year of life. The Feeding Infants and Toddlers Study, a cross-sectional study of US children aged 4 to 24 months conducted in 2002, documented infant and child food intake on the basis of one or two 24-hour recalls but did not collect detailed information on younger infants.5 In the study described here, we attempted to fill this knowledge gap by assessing milk and formula consumption among a large sample of US infants, as well as changes in their diet throughout their first year of life as they were introduced to complementary foods.
We analyzed data from the Infant Feeding Practices Study II (IFPS II), a longitudinal mail survey of mothers of infants conducted by the US Food and Drug Administration in 2005-2007. The sample was drawn from a consumer-opinion mail panel distributed throughout the United States. Participants in the panel were asked whether any household member was currently pregnant. Panel members identified as being pregnant were then invited to participate in the study, with the time of their enrollment based on their expected due date. IFPS II participants completed mail questionnaires almost monthly from the third trimester of their pregnancy through their infant's first year of life. Additional details about the study methodology can be found elsewhere.6
Postpartum questionnaires were mailed to participants when their infants were ∼1, 2, 3, 4, 5, 6, 7½, 9, 10½, and 12 months of age. Each questionnaire included a food-frequency table on which mothers were asked to indicate how often their infants were fed various types of food or drink in the previous 7 days (the 7 days before completing the questionnaire). Mothers could respond in times per week or times per day. With the exception of the month 1 (neonatal) questionnaire, the food-frequency table on each questionnaire consisted of 18 categories of foods and drinks. The table in the neonatal questionnaire consisted of 6 of the food/drink categories in the later questionnaires plus water and sugar water. The neonatal questionnaire also asked mothers who had ever breastfed about in-hospital feeding of infant formula, water, and sugar water.
Although most mothers completed each survey within several days of receiving it, many did not do so until several weeks or even months later. Therefore, we analyzed the feeding data according to the age of the infant when the questionnaire was actually completed. We created the following age windows for analysis: 3 to 6 weeks (∼1 month), 7 to 10 weeks (∼2 months), 11 to 14 weeks (∼3 months), 15 to 18 weeks (∼4 months), 19 to 23 weeks (∼5 months), 24 to 28 weeks (∼6 months), 29 to 35 weeks (∼7½ months), 36 to 42 weeks (∼9 months), 43 to 50 weeks (∼101/2 months), and 51 to 59 weeks (∼1 year). For example, if a mother completed the 6-month questionnaire when her infant was 37 weeks old, we incorporated data from that questionnaire with data for infants aged 36 to 42 weeks instead of with data for infants aged 24 to 28 weeks. If a mother completed 2 questionnaires while her infant was in the same age window, we used the data from the first questionnaire she completed.
Breast Milk Consumption
We divided the infants into 5 mutually exclusive categories of breast milk consumption on the basis of their mothers' reports of the percentage of milk or formula feedings in the previous 7 days in which their infants were given breast milk: all breast milk, more than two thirds breast milk, one third to two thirds breast milk, less than one third breast milk, and no breast milk. Because in-hospital feeding data did not specify the number of feedings that infants received, we used only 3 categories for this time period: breast milk only, mixed breast milk and formula, and formula only.
We determined the percentage of infants who consumed each type of food or drink at least once in the previous 7 days and also the percentage who consumed at least 1 food or drink in each of the following categories: (1) breast milk or formula; (2) water or sugar water; (3) cereal; (4) fruits and vegetables; (5) meat or meat substitute; (6) other milk products; and (7) fatty or sugary foods.
To describe the variety of foods in the infants' diet, we classified each reported feeding according to the type of solid foods or liquids they consumed. We considered infants to have consumed solid foods in 1 of 3 groups if they were reported to have consumed foods in that group an average of at least once per day. For example, if an infant ate fish once per week and eggs 6 times per week, it would count as consuming food in the meat group once per day. The 3 solid food groups were meat or meat substitutes (meat/chicken, fish, peanuts, eggs, soy foods), fruits or vegetables (not including juice), and cereals (including infant cereal and other cereals/starches). We then created the following 8 categories of food consumption: (1) 3 groups of solids consumed at least once per day; (2) 2 groups of solids consumed at least once per day; (3) 1 group of solids consumed at least once per day; (4) some solids consumed but no group consumed at least once per day; (5) liquids other than breast milk or formula but no solids consumed; (6) only breast milk and formula consumed; (7) only formula consumed; and (8) only breast milk consumed.
Predictors of Selected Infant Feeding Practices
To identify factors associated with various infant feeding practices, we selected 5 key indicators of adherence to infant feeding recommendations: (1) breastfeeding in the hospital; (2) not consuming any solid foods at 4 months; (3) breastfeeding at 6 months; (4) consuming a varied diet with at least daily consumption of cereals, meats or meat substitutes, and fruits or vegetables at 9 months; and (5) not consuming fatty or sugary foods at 1 year. We then calculated the overall percentage of infants whose consumption met each of these criteria, as well as the percentage within various demographic categories defined by mother's age, parity, education, family income, participation in the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), race, and region of residence. We also examined associations between the infants' feeding practices at each age interval and their feeding practices at earlier age intervals. For example, we assessed the extent to which consumption of a varied diet at 9 months was associated with consumption of solids at 4 months and with consumption of breast milk at 6 months. We used SAS 9.1 (SAS Institute, Inc, Cary, NC) to examine these associations, and we used logistic regression to predict the likelihood of various infant feeding practices on the basis of a model that accounted for all maternal sociodemographic characteristics and earlier infant feeding practices. The sample size for the logistic regressions was limited to the observations with complete data on sociodemographic characteristics and earlier feeding practices.
Infants' Consumption of Liquids and Solids
Breast Milk or Formula
Although 83% of infants in this study were breastfed in the hospital, the diets of 42% of these infants were supplemented with formula (Fig 1). By 3 months, the percentage of infants who received formula (61%) roughly equaled the percentage who received breast milk. The percentage of infants who received both breast milk and other milk or formula declined continuously from 35% in the hospital to only 14% at 9 months but then rose again as infants began to consume cow's milk. The percentage of breastfed infants whose diets were supplemented with infant formula declined from 42% in the hospital to 34% at 3 months of age and remained at this level until 9 months of age. More than half of the infants classified as “mixed-milk feeders” received more than two thirds of their feedings from breast milk. Infant formula accounted for most of the “other-milk” feedings of infants up to the age of 10½ months, but by the age of 1 year, 81% of the infants had begun consuming cow's milk (Table 1).
Infants' consumption of water and sugar water was only assessed in the neonatal questionnaire. Responses to this questionnaire showed that 13% of the infants were given sugar water while in the hospital and 10% of the infants were receiving water without sugar at the age of 1 month despite findings that infants who do not consume solid foods have no need for solute-free water.7,8
By 3 months of age 18% of the infants were consuming infant cereal, and by 4 months of age 40% were consuming infant cereal. Infants were introduced to infant cereal at a median age of slightly older than 4 months and to other cereals at a median age of ∼8 months. By the end of infancy, consumption of infant cereal had dropped off markedly, with only 46% of the infants still consuming infant cereal at 1 year (Table 1).
Fruits and Vegetables
Fruits and vegetables were introduced into the infants' diet at a median age of between 5 and 6 months, and by 7½2 months, >90% of the infants were consuming fruits and vegetables. The percentage of infants who consumed 100% fruit juice rose continuously throughout infancy, from <20% at 5 months to 77% at 1 year (Table 1).
Meat and Meat Substitutes
The median age of infants when introduced to meat or combination foods containing meat was ∼8 months, and by the age of 1 year, nearly 97% of all the infants were consuming meat or meat substitutes, including 60% who were consuming eggs, 25% who were consuming peanuts or peanut butter, 18% who were consuming fish or shellfish, and 6% who were consuming soy products (Table 1).
Cow's Milk and Milk products
At 10½ months, only 17% of the infants were consuming cow's milk, but by 12 months 81% of them were. Other dairy foods, such as cheese and yogurt, were introduced to infants at a median age of ∼10 months.
Fatty or Sugary Foods
By 1 year of age, approximately half of the infants were consuming French fries and candy, cookies, or cake, although only 15% were consuming sweetened drinks such as soda or juice drinks.
Type of Feeding According to Age
From birth through 3 months of age, at least 80% of the infants were fed nothing but breast milk or formula;between 4 and 5 months, most of the infants were introduced to solid foods;and from 6 months onward, >80% of the infants consumed solid foods on a daily basis (Fig 2). At 6 months, the majority of infants consumed at least 1 serving per day of solid foods only from 2 food groups, but by 9 months, most of them consumed at least 1 serving per day from all 3 food groups. Except in the hospital, only a small percentage of the infants consumed liquids other than breast milk or formula without also consuming solid foods.
Predictors of Infant Feeding Behavior
Predictors of In-Hospital Breastfeeding
Initiation of breastfeeding in the hospital was positively associated with living in the western region of the country, Hispanic ethnicity, and mothers having had at least some college education but negatively associated with WIC enrollment (Table 2). Although mothers >30 years old had a higher unadjusted rate of breastfeeding initiation than those who were ≤30 years old, older maternal age was actually negatively associated with breastfeeding initiation after we controlled for other characteristics.
Predictors of not Eating Solid Foods at 4 Months
Overall, 41% of the infants were consuming solid foods at 4 months of age (Table 2). Adherence to the recommendation to not introduce solid foods before their infants were 4 months was more common among college educated mothers, mothers >25 years old, and those living in the western region of the country and less common among WIC participants. Although the mothers with higher income seemed to be more likely to follow this recommendation, the association between income and adherence to this recommendation was actually negative after we adjusted for other characteristics. The infants who were initially breastfed were far more likely to have solids introduced after 4 months than those who were fed only infant formula in the hospital.
Predictors of Continued Breastfeeding at 6 Months
Although primiparous mothers were more likely than multiparous mothers to initiate breastfeeding, they were less likely to still be breastfeeding when their infants were 6 months of age (Table 2). Also, although the rates of breastfeeding at 6 months were higher among college-educated mothers and those who were not WIC participants, family income was actually negatively associated with breastfeeding rates at 6 months after we controlled for these 2 factors. We also found that Hispanic mothers were more likely to initiate breastfeeding but were less likely to continue it for 6 months than white mothers, that mothers living in the western region of the country had the highest rate of breastfeeding initiation and the highest rate of breastfeeding at 6 months, and that mothers who began feeding their infants solids by 4 months were far less likely than those who did not to be breastfeeding when their infants were aged 6 months.
Predictors of a Varied Diet at 9 Months
By the age of 9 months, 45% of the infants consumed at least 1 serving per day of food from each of 3 food groups (cereals, fruits and vegetables, and meats or meat substitutes) (Table 2). The introduction of solids by 4 months was the only factor we found to be significantly associated with the infants' consumption of a varied diet at 9 months.
Predictors of not Consuming Fatty or Sugary Foods at 1 Year of Age
At the age of 1 year, only one third of the infants were reported to have refrained from sweetened drinks, French fries, cookies, cakes, or candies in the previous week (Table 2). Refraining from fatty or surgary foods was less common among infants with at least 1 older sibling, more common among Hispanic infants than among white or black infants, and less common among those introduced to solid food by 4 months of age than among those who were not.
In this study, we documented the dramatic transitions in dietary consumption that occurred among infants during their first year. The transition from a diet of virtually nothing but breast milk, infant formula, or both to a varied diet of foods from all food groups began for most infants at ∼4 to 5 months of age and continued throughout their first 12 months. Infant cereal was usually the first food other than milk or formula given to infants and remained the most common supplementary food until infants were ∼8 months of age. Fruits and vegetables were introduced at a median age of 5 to 6 months, and meats were introduced at a median age of ∼8 months. By 1 year of age, more than half of the infants were consuming a diet that included not only cereals, fruits, vegetables, meats, and milk products but also foods high in sugar or fat but low in nutrient density.
Several of the infant feeding patterns that we found are cause for concern. For example, we found that the diets of almost half of the breastfed infants were supplemented with infant formula while the infants were still in the hospital. Although one might expect breastfeeding women to turn to infant formula as a supplement to breast milk when they return to work or other activities, there is no medical reason for most infants to be given formula while they are in the hospital.9 The AAP recommends that breastfed infants be given formula in the hospital only when medically necessary, because early supplementation can lead to later problems with breastfeeding.2
The AAP Section on Breastfeeding and the Committee on Nutrition agree that introduction of solid foods should not occur any earlier than 4 months of age. In this study, we found that 41% of the infants were already consuming solid foods at 4 months of age, the beginning of the window recommended by the AAP Committee on Nutrition1 and well before the age for introducing solid food recommended by the AAP Section on Breastfeeding.2 Not only do infants not need solid foods before 4 months,1 but the consumption of solid foods exposes infants to additional pathogens and solid foods do not provide the immunologic benefits of the breast milk they replace.2 We demonstrated here that early introduction to solid foods is also a risk factor for both earlier cessation of breastfeeding and increased consumption of fatty or sugary foods at 1 year of age. Our finding that so many infants were consuming solid foods at such a young age indicates that physicians may need to provide clearer guidance to parents about when to introduce their infants to solid food and the risks of starting too early.
Although we found that cereal was typically introduced into the infants' diet quite early, most infants were not introduced to meat until they were ∼8 months of age. Because fetal iron and zinc stores are typically depleted by ∼ 6 months of age,10 good sources of bioavailable iron and zinc need to be among the first solid foods given to breastfed infants. Unfortunately, the bioavailability of the iron compounds commonly used in infant cereal is relatively low,11 and although some infant cereals are now fortified with zinc in addition to iron, the bioavailability of the zinc in these cereals has not been reported. Early introduction of meat has the advantage of providing a good source of iron and zinc in a highly bioavailable form,1,12,13 and pureed meats have been shown to be well tolerated by infants as a first complementary food.14
Approximately two thirds of the infants in our study were consuming high-fat/high-sugar foods at 1 year of age. These foods provide high amounts of calories but are poor in nutrient content. The nutritional requirements of older infants leave little room in their diet for such foods.3 Given the increasing prevalence of obesity among children and infants,15 it is important that infants establish healthy eating behaviors as early as possible. Although children's eating patterns can change over time, researchers have shown that healthy eating patterns begun early in life form a foundation that may continue for many years16; unfortunately, the same has been shown to be true of unhealthy eating patterns.3 Counseling parents of infants about the importance of good nutrition could help increase the number of infants who establish healthful eating habits and, thus, reduce their risk for obesity and other chronic diseases that have been linked to diet.
This study suffers from some key limitations. Although the sample was drawn from across the country, minority groups were underrepresented, the educational level of participants was higher than the national average,6 and study participants had to be English-speaking and have a literacy level high enough to complete the questionnaires. In addition, only mothers who were willing to complete a series of monthly questionnaires were included, creating an unknown bias in the sample. The sample at the end of the study was further biased in favor of women with higher socioeconomic status, because those who dropped out of the study between the neonatal questionnaire and the 12-month questionnaire were more likely to be younger, to be less educated, to have lower income, to be participating in WIC, to be nonwhite, and to live in the southern region. In addition, we were unable to estimate the nutrient intake of the infants (because the IFPS II data on infant feeding did not include information on serving size) or to precisely characterize the infants' diets (because of the broad food categories asked about in the IFPS II questionnaires).
However, the study also has some key strengths. To our knowledge, the IFPS II was the largest study of infant feeding conducted in the United States, with >3000 infants sampled in the neonatal period and nearly 1800 followed through 12 months of age. Although an earlier study conducted beginning in 1992 collected detailed longitudinal information about the diets of infants beginning at 2 months of age, only 98 infants were followed.17 The National Health and Nutrition Examination Survey, which is used to monitor the diets of all Americans, includes only a cross-sectional sample of ∼500 infants every 2 years. Because the IFPS II followed infants longitudinally from birth through 1 year of age with questionnaires sent to their mothers nearly every month, we were able to characterize changes in infants' diets on a nearly monthly basis and to assess associations between feeding behaviors at 1 time during infancy with feeding behaviors at later times. In addition, because the IFPS II asked mothers about feeding over the previous 7 days, it minimized the risk for recall error that is typically a problem with retrospective reporting of infant feeding other than breastfeeding.18
In this study, we identified several infant feeding practices of concern, including substantial formula supplementation in the hospital, early introduction of solid foods, late introduction of meats, and feeding of high-fat/high-sugar foods to infants. Because of their frequent contact with infants and their parents, clinicians have a unique opportunity to advise new parents about recommended infant feeding practices. By being aware of these infant feeding recommendations and communicating them to parents, clinicians can help start children on the road to a healthy lifestyle.
This study was funded by the Food and Drug Administration, Centers for Disease Control and Prevention, Office of Women's Health, National Institutes of Health, and Maternal and Child Health Bureau in the US Department of Health and Human Services.
- Accepted June 4, 2008.
- Address correspondence to Laurence M. Grummer-Strawn, PhD, Centers for Disease Control and Prevention, National Centerfor Chronic Disease Prevention and Health Promotion, Division of Nutrition, Physical Activity, and Obesity, 4770 Buford Hwy, MS K25, Atlanta, GA 30341. E-mail:
The authors have indicated they have no financial relationships relevant to this article todisclose.
The findings and conclusions in this article are those ofthe authorsand do not necessarily represent the official position of the Centersfor Disease Control and Prevention or the Food and Drug Administration.
- 1.↵American Academy of Pediatrics, Committee on Nutrition. Pediatric Nutrition Handbook. Kleinman RE, ed. 6th ed. Elk Grove Village, IL: American Academy of Pediatrics; 2008: In press
- Gartner LM,
- Morton J,
- Lawrence RA, et al
- Fein SB,
- Labiner-Wolfe J,
- Shealy KR,
- Li R,
- Chen J,
- Grummer-Strawn LM
- 9.↵Academy of Breastfeeding, Medicine Protocol Committee. Clinical Protocol Number 3: Hospital Guidelines for the Use of Supplementary Feedings in the Healthy Term Breastfed Neonate. Academy of Breastfeeding Medicine, Inc; 2002. Available at: www.bfmed.org/ace-images/ProtocolSuppl3rev.pdf. Accessed July 29, 2008
- 10.↵World Health Organization. Nutrient Adequacy of Exclusive Breastfeeding for the Term Infant During the First Six Months of Life. Geneva, Switzerland: World Health Organization; 2002
- Davidsson L,
- Kastenmayer P,
- Szajewska H,
- Hurrell RF,
- Barclay D
- Krebs NF,
- Hambidge KM
- Copyright © 2008 by the American Academy of Pediatrics