Infant Metabolic Alkalosis and Soy-Based Formula -- United States
Three cases of a Bartter-like syndrome in infants were reported to CDC from Memphis, Tennessee, on July 26, 1979. The infants were less than 10 months of age and were failing to gain weight. They had poor appetites, and one had a history of constipation. All were hypochloremic and hypokalemic, with varying degrees of alkalosis and microhematuria. The 3 infants were taking the same brand of soy-based formula.
To further investigate this possible association, CDC surveyed a sample of pediatric nephrologists throughout the country for cases of metabolic alkalosis diagnosed since January 1, 1979, in infants with a history of failure to thrive, anorexia, or constipation. Infants known to have pyloric stenosis, cystic fibrosis, or diuretic therapy were excluded.
An additional 15 cases were ascertained through the survey, and another 16 cases were determined from other sources. Cases were scattered throughout the country. The infants ranged in age from 2 to 9 months; none died. There was no unusual sex distribution.
Feeding history was available in 27 of the 31 cases. Of these, 26 were on Neo-Mull-Soy (Syntex, Palo Alto, California), the same formula used by the 3 index cases. Neo-Mull-Soy represents 10%-12% of the soy-based formula market. After diagnosis of the alkalosis, infants who were placed on chloride supplement responded favorably; those who, after treatment for and recovery from the alkalosis, went back on the formula--but without chloride supplementation--had a recurrence.
The manufacturer of Neo-Mull-Soy has voluntarily stopped manufacturing this product, halted its distribution to wholesalers, and requested that wholesalers stop sales to retailers. Syntex has also issued a mailgram to pediatricians and pediatric residents notifying them of the problem.
Reported by: JS Levy, MD, Memphis-Shelby County Health Dept, Memphis, Tennessee; S Roy, MD, Memphis; RH Hutcheson Jr, MD, State Epidemiologist, Tennessee State Dept of Public Health; AB Gruskin, MD, Philadelphia, Pennsylvania; S Hellerstein, MD, Kansas City, Missouri; M Linshaw, MD, Kansas City, Kansas; S Alexander, MD, JD Liberti, MD, Portland, Oregon; H Harrison, MD, Louisville, Kentucky; G Lum, MD, Denver, Colorado; LJ Cunningham, MD, Galveston, Texas; EH Garin, MD, Gainesville, Florida; Div of Nutrition, Bur of Foods, Food and Drug Administration; Birth Defects Br, Chronic Diseases Div, Bur of Epidemiology, CDC.
Editorial Note: Bartter syndrome is characterized by hypochloremic, hypokalemic alkalosis; normal blood pressure; and increased serum levels of renin and aldosterone. The onset is usually during the first year of life. The pathogenesis is not known.
The high percentage of affected infants on Neo-Mull-Soy formula and the fact that infants who were switched to other soy formulas did not have recurrence both support the casual association between Neo-Mull-Soy formula and this outbreak.
Insufficient intake of chloride is a known cause of metabolic alkalosis. The cause of this outbreak is not yet clear, but it is possible that the chloride concentration in this formula falls below the daily requirement for infants, if they are not also receiving chloride from other dietary sources. The current tendencies to delay the addition of solids to infants' diets and to remove sodium chloride from commercial and home-prepared baby foods might be additional contributing factors.
There are no regulations pertaining to the optimal level of chloride in infant formulas. The Committee on Nutrition of the American Academy of Pediatrics recommends a minimum of 11 milliequivalents per liter in infant formula (1).
1. Committee on Nutrition, American Academy of Pediatrics: Commentary on breast-feeding and infant formulas, including proposed standards for formula. Pediatrics 57:278-285, 1976. Editorial Note--1996: At the time of this cluster of cases of hypochloremic metabolic alkalosis, infant formula was regulated under 21 CFR 105.65, Infant Foods. This regulation specified minimum levels of certain nutrients for infant formulas, including protein, fat, and some vitamins and minerals; a level for chloride was not specified. If the specified levels of nutrients were not present in the formula, the label was required to state that the diet should be supplemented. The incident described in this report prompted the Infant Formula Act of 1980*--the amendment of the federal Food, Drug, and Cosmetic Act that established a new section 412 (21 U.S.C. 350a) and created a separate category of food designated as infant formula. Section 412 requires that infant formulas meet specified standards of quality and safety and contain all required nutrients, including chloride, at specified levels. The Infant Formula Act of 1980 was the first in a series of major legislative and regulatory steps taken to ensure the safety of infant formulas** (1,2).
This episode underscores the need for regular and adequate testing of infant formulas. Several events may have contributed to the formula chloride deficiency, including removal of sodium chloride from the formula for the purpose of reducing the sodium content of infant diets. The cummulative effect of these contributing events led to a deficiency that was not recognized because regular testing for chloride content was not conducted.
In follow-up to the investigation in 1979, CDC established a registry of children who developed hypochloremic metabolic alkalosis following consumption of chloride-deficient Neo-Mull-Soy and Cho-Free, another soy-based formula manufactured by Syntex. Based on these data, the National Institutes of Health conducted a follow-up study to determine whether the risk for developmental delays or deficiencies was increased in these children (3). The study determined that by age 9-10 years, the children appeared to have recovered from their early growth failure and to have achieved normal cognitive development. However, these children remained at potential risk for deficits in language skills that require expressive language abilities (3).
This investigation highlights the critical importance of developing and using appropriate case definitions for surveillance and in investigations of outbreaks of both infectious and noninfectious origin. The original diagnosis of these cases was Bartter syndrome, a condition that causes metabolic alkalosis from renal loss of potassium and requires a large replacement dose of potassium chloride throughout life to maintain metabolic homeostasis. The children who had hypochloremic metabolic alkalosis as the result of consuming chloride-deficient formula quickly recovered following treatment with small doses of potassium chloride. This clinical response provided a clue to the physician who reported the first three cases that the formula might be the cause of the metabolic alkalosis. As a result, CDC's survey of pediatric nephrologists was used to search for cases of metabolic alkalosis resembling Bartter syndrome, rather than confirmed cases of that condition. If the case definition in this survey had been restricted to Bartter syndrome only, the association may not have been detected.
The outbreak described in this report highlights the value of a rapid response capability for local and state health departments and the Public Health Service and the important role played by clinicians in identifying public health emergencies. The sequence of problem recognition, investigation, and response unfolded rapidly: on July 26, 1979, CDC was notified of the three cases from Memphis and of the causal hypothesis related to infant formula as suggested by the attending physician. On July 27, two of CDC's Epidemic Intelligence Service (EIS) officers reported for their first day of work on assignment to CDC's Birth Defects Branch and assisted in developing a strategy for collecting information about feeding histories of children with metabolic alkalosis. On July 30, the nationwide survey of pediatric nephrologists was conducted. On August 1, one EIS officer traveled to the manufacturer's corporate headquarters to meet with company officials and three pediatricians. The company tested several formula batches before the meeting and found that none contained sufficient chloride. On August 2, after meeting with representatives of the Food and Drug Administration, the company halted manufacture of the formulas, initiated a voluntary recall of the products, and notified health-care professionals throughout the country about the problem. The MMWR article describing the occurrence was released to the news media that same day, only 7 days after CDC received notification of the first three cases from Memphis.
1996 Editorial Note by: Shane Roy, III, Dept of Pediatrics, Univ of Tennessee, Memphis. Frank Greenberg, National Center for Human Genome Research, National Institutes of Health. Gillian Robert-Baldo, Nicholas Duy, John Wallingford, Office of Special Nutritionals, Center for Food Safety and Applied Nutrition, Food and Drug Administration. Heinz Berendes, Div of Epidemiology, Statistics and Prevention Research, National Institute for Child Health and Development, National Institutes of Health. J David Erickson, DDS, Birth Defects and Genetic Diseases Br, Div of Birth Defects and Developmental Disabilities, National Center for Environmental Health; Jose Cordero, MD, National Immunization Program, CDC.
1. Infant Formula Quality Control Procedures (47 FR 17016, April 20, 1982); Enforcement Policy; Infant Formula Recalls (47 FR 18832, April 30, 1982); Infant Formula; Labeling Requirements (50 FR 1833, January 4, 1985); Nutrient Requirements for Infant Formula (50 FR 45106, October 30, 1985); Exempt Infant Formula (50 FR 48183, November 22, 1985); Infant Formula Recall Requirements (54 FR 4006, January 27, 1989); and Infant Formula Record and Record Retention Requirements (56 FR 66566, December 24, 1991).
2. Current Good Manufacturing Practice, Quality Control Procedures, Quality Factors, Notification Requirements and Records and Reports, for the Production of Infant Formula (61 FR 36154) (Proposed Rule).
3. Malloy MH, Graubard B, Moss H, et al. Hypochloremic metabolic alkalosis from ingestion of a chloride-deficient infant formula: outcome 9 and 10 years later. Pediatrics 1991;87:811-22.
* Public law 96-359.
** Public law 99-570.