Fluoride content of infant formulas: soy-based formulas as a potential factor in dental fluorosis.

Recent reports of art increased prevalence of dental fluorosis in fluoridated and nonfluoridated communities have led to a reassessment of the amount offluoride (F) being ingested infants and young children. Manufacturers of milk-based formulas have taken steps to reduce the F concentration to negligible levels. Reduction of F concentration in soy-based formulas is more difficult because ofF binding to phytate and tricalcium phosphate. This study assessed the F content of 3 milk-based and 4 soy-based infant formulas. Three types of preparations in each brand were tested: ready-to-feed, liquid concentrate, and powdered concentrate. Concentrates were diluted according to the manufacturers’ recommendations with water containing various concentrations of F. The ionic F was measured using an ion-specific electrode. Acid-diffusable F was measured after separation by the Taves method. Soy-based formulas contained higher levels ofF than their milk-based counterparts. In the case of ready-to-feed and liquid preparations the differences were statistically significant. The mean F concentrations for soy-based products were: ready-to-feed, 0.30 mg/l; liquid concentrate diluted with deionized water, 0.24 m g/l ; and powdered concentrate dilu ted with deionized water, 0.08 mg/l. Both soy-based ready-to-feed and diluted liquid concentrate formulas provide daily F dosages which, in combination with supplemental dietary F (0.25 rag~day), exceed the currently defined norms for optimum daily F intake. Recent reports have described an increased prevalence of dental fluorosis in children’s teeth in both fluoridated and nonfluoridated communities (Aasenden and Peebles 1974; Forsman 1977; Messer and Walton 1980; Leverett 1982,1986) greater than that reported in the early studies of Dean et al. (1942). Soparkar and DePaola (1985) reported finding an unexpectedly high number of school children with dental fluorosis residing in nonfluoridated areas of Massachusetts. Leverett and Levy (1983) reported similar findings in children both fluoridated and nonfluoridated communities in an earlier study. Horowitz et al. (1984) speculated that the increased prevalence of dental fluorosis in younger children could be related to infant formulas and other foodstuffs processed with fluoridated water. Studies have indicated that in several species, including humans, a daily fluoride (F) intake of 0.1 mg/kg body weight during the period of enamel calcification is sufficient to cause mild dental fluorosis (Forsman 1977; Suttie et al. 1972). It should be emphasized that the levels of fluorosis reported by these investigators are in the very mild and mild categories described by Dean et alo (1942) and are, for the most part, not discernable by the layman. Prior to 1978, Adair and Wei (1978), Stamm and Kuo (1977), and Singer and Ophaug (1979) found great variability in the F content of infant formulas. The concentrations ranged from 0.08 to 0.78 mg/1 F for ready-tofeed milk-based formulas and from 0.31 to 0.92 mg/1 in ready-to-feed soy-based formulas. Changes in the manufacturing process have subsequently led to a reduction in the amount of F in milk-based formulas to a considerably lower level. Because soy-based infant formulas contain phytates and tricalcium phosphates, both of which bind F, the potential for larger than optimum dosage of F for infants using soy-based formulas still exists. A recently published study which evaluated the F content of infant formulas purchased in various geographic regions of the United States (Johnson and Pediatric Dentistry: September, 1988 N Volume 10, Number 3 189 Bawden 1987) indicates that the F concentration of infant formulas has been reduced from that reported for some products prior to 1980. This study reported higher F concentrations in all the groups of soy-based formulas that were tested compared to the milk-based formulas tested. Several studies have evaluated the dietary F intake of infants (Stamm and Kuo 1977; Adair and Wei 1978; Singer and Ophaug 1979; Ophaug et al. 1985). Most recently, Ophaug et al. (1985) reported findings based on the analysis of market basket food collections in 4 dietary regions of the United States. Their data indicated that the average dietary F intake of 6-monthold infants did not exceed 0.52 mg/kg, however. This study did not include soy-based formulas in the composite groups analyzed. Recent surveys conducted at the Eastman Dental Center of parents of 6and 7-year-old children indicate that approximately 15% of the children were fed soybased formula as infants. The current philosophy regarding choice of formula for infants seems to support changing an infant to soy-based formula at the first indication of any problem with feeding (Hill et al. 1984; Gillooly et al. 1984; American Academy of Pediatrics Committee on Nutrition 1983). If a familial history of food allergies exists, infants sometimes are started out on soy-based formulas. In light of this information and as a part of a larger study evaluating the prevalence of, and possible explanations for, dental fluorosis, the aim of the present study was to evaluate and compare the F content of several milk-based and soy-based infant formulas. Materials and Methods Three types of milk-based and soy-based infant formulas were analyzed: ready-to-feed -canned liquid formula ready for direct feeding without dilution; diluted liquid concentrate -canned liquid formula diluted 1:1 with water prior to feeding; and diluted powder concentrate -dry formula prepared with a standard dilution of 1 scoop powder for each 2 ounces of water. The formulas were prepared according to manufacturers’ recommendations prior to analysis and were diluted with water containing various concentrations of F, depending on the purpose of the analysis. The six types of formulas tested were: Ready-to-feed, milk base (RF-M) Ready-to-feed, soy base (RF-S) Liquid concentrate, milk base (LC-M) Liquid concentrate, soy base (LC-S) Powder, milk base (P-M) Powder, soy base (P-S). The specific brand of each type of formula was selected on the basis of its availability in a variety of grocery stores and pharmacies in the Rochester, New York area. The brands of milk-based formula tested were: RF-M -Similac ® (3 samples), SMA + iron ® (2 samples), and Enfamil® (3 samples); LC-M: SMA + iron (2 samples) and Enfamil (3 samples); P-M: Enfamil samples). The brands of soy-based formulas tested were: RFS: Prosobee® (6 samples), Nutrimigen® (4 samples), and Isomil® (6 samples); LC-S: Prosobee (5 samples), Isomil (3 samples), and Nursoy® (3 samples); P-S: Prosobee samples) and Isomil (3 samples). Fluoride was separated from 3 ml of the prepared infant formulas as hydrofluoric acid (HF), appropriately buffered, and analyzed directly using a fluoride combination electrode (Model 96-09 -Orion Research Inc; Orion, MA). The Taves (1968) separation method was employed to separate the acid-diffusable F from the sample. A calibration curve was constructed of standards (dilutions from NaF primary standards) and the reagent blank, all of which had undergone sample separation. The reagent blank was low, 0.02 ~tg F (+ .005 SE), and was subtracted from all standards and unknowns. Each of the samples was analyzed in triplicate, with results presented as mg F/1 of sample (weight/volume). Student’s t-test was used to determine statistical significance of the findings.