Infant formula preparation, handling, and related practices in the United States.

Although infants in the United States are commonly fed formula, the details of how mothers prepare and handle formula are not well described in the literature. Widely available recommendations (1-5) state that caregivers should not dilute or concentrate formula; mix formula with warm tap water; use unsterilized bottles, nipples, or water; keep prepared formula at room temperature; heat bottles in a microwave oven; put the baby to bed with a bottle; or add cereal or sweeteners to bottles of formula.

Extensively altering the formula mix constitutes a public health issue because diluting or concentrating formula increases an infant's vulnerability to hyponatremic seizures if a large amount of additional water is fed (6,7) or to hypernatremic dehydration if too little water is used (6-8). These consequences can cause permanent disability or death and often require hospitalization (8-12). Although diluting or concentrating formula may be common in developed countries (1315), the occurrence of severe consequences from these practices appears to be relatively infrequent.

Reconstituting formula with warm tap water is one of several practices that may increase the amount of lead in prepared formula (2,3). Infants who are fed reconstituted formula ingest a large quantity of water per unit of body weight, so that any lead in the water used to reconstitute formula is a concern (1,16-20).

Whether equipment and water should be sterilized is controversial. Although consumer education materials recommend sterilizing (2-5,21,22), a review by Schuman (23) concluded that sterilization is not necessary when water that has safe levels of bacteria is available. Also, physicians concerned about lead intoxication recommend not boiling water used to mix formula because boiling concentrates lead (24). Since this discussion, however, the discovery of the parasite Cryptosporidium in some public water systems has caused outbreaks of illness, raising concern about using water that has not been boiled (25). The importance of another formula-handling practice, keeping prepared formula cold, is not disputed.

Heating bottles in the microwave is not recommended because of possible scalding or exploding bottles (1), and infant burns severe enough to require hospitalization and even amputations have been reported (26,27). A protocol for safe microwave heating of bottles has been developed because the authors believe the practice will remain common (28).

Putting an infant to bed with a bottle containing a liquid high in carbohydrate can cause dental caries in a pattern called baby bottle tooth decay (29-32), which is sometimes so extensive that extraction of teeth is required in children as young as 18 months (1).

Adding cereal or sweeteners to bottles is not recommended. However, even though infants seem to regulate energy intake so that cereal displaces the more nutrient-dense formula (33), early feeding of solid food appears to be tolerated by most infants (34-36) and does not seem to affect growth (33,37). In some infants, feeding solids early increases coughing (34,38) and eczema (34). Feeding sweeteners to infants is not recommended; honey has been linked to infant botulism (1,39-41) and other sweeteners provide only empty energy.

This study describes the extent to which mothers follow the recommended formula feeding practices and the relation between compliance with the practices and receipt of instruction in formula preparation from a health care professional, demographic characteristics, and breast-feeding status (because breast-feeding mothers are more likely to follow certain infant feeding recommendations [42-50]).

METHODS

We analyzed data from the US Food and Drug Administration's Infant Feeding Practices Study (IFPS), an unpublished longitudinal survey of mother-infant pairs.1 Mothers were a subsample of a national mail panel of about 500,000 house-holds. Panel members were screened for pregnancy, and the 3,155 women who were listed as being in their third trimester of pregnancy were asked to participate. Data were collected. between February 1993 and October 1994. Respondents were considered ineligible if they were not pregnant (n=297), it' mother or infant had medical problems (n=108), if the birth was not singleton (n=22), or for certain administrative reasons (eg, baby was born too early for the questionnaire to be administered on time (n=113). Nonrespondents were defined as those who failed to return any 1 of the first 3 questionnaires. The study response rate was 69% (1,803 of the 2,615 respondents not deemed ineligible). Most of the sample loss occurred with the first questionnaire.

Data were collected prenatally, shortly after birth, and 9 additional times - at infant ages 1,2,3,4,5,6,7, 9, and 12 months. The questionnaires sent in months 2 to 12 were divided into 8 modules. Three core modules were administered each month, but the other 5 modules were only included some of the time. The data used here are primarily from the 3 infant …