Ingestion of aspartame, 50 mg/kg, by normal lactating women results in a small, but statistically significant, rise in overall aspartate and phenylalanine milk concentrations (9). Milk aspartate levels rose from 2.3 to 4.8 μmol/dL during a 4-hour fasting interval after the aspartame dose. The rise in phenylalanine milk levels during the same interval was approximately 0.5-2.3 μmol/dL. The investigators of this study concluded that these changes, even if spread over an entire 24-hour period, would have no effect on a nursing infant's phenylalanine intake (9). A 2015 study of seven women that self-reported aspartame consumption within the previous 24 hours found that aspartame was not detected in breast milk (29). Phenylalanine and aspartate levels were not measured.

However, mothers or infants with PKU need to carefully monitor their intake of phenylalanine. In 2001, the American Academy of Pediatrics classified aspartame as an agent to be used with caution during breastfeeding if the mother or the infant has PKU (30).

Aspartame is a nutritive sweetening agent used in foods and beverages. It contains 4 kcal/g and is about 180-200 times as sweet as sucrose. The product was discovered in 1965 and obtained final approval by the FDA as a food additive in certain dry foods in 1981 and in carbonated beverages in 1983 (1). Aspartame is probably the most extensively studied food additive ever approved by the FDA (2,3). Chemically, the compound is N-L-α-aspartyl-L-phenylalanine, 1-methyl ester (i.e., the methyl ester of the amino acids L-phenylalanine and L-aspartic acid).

Human Data: Aspartame is broken down in the lumen of the gut to methanol, aspartate, and phenylalanine (1, 2, 3, 4). The major decomposition product when the parent compound is exposed to high temperatures or in liquids is aspartyl-phenylalanine diketopiperazine (DKP), a product formed by many dipeptides (3, 4, 5). The rate of conversion of aspartame to degradation products depends on pH and temperature (3). In addition to methanol, the two amino acids, and DKP, other degradation products are L,L-β-aspartame (aspartame is commercially available as the L,L-α isomer), three dipeptides (α-Asp-Phe, β-Asp-Phe, and Phe-Asp), and phenylalanine methyl ester (3). The dipeptide compounds are hydrolyzed to the individual amino acids in the gut, and these, along with methanol, will be evaluated in later sections. None of the other degradation products have shown toxic effects after extensive study (3). Moreover, the DKP compound and the β isomer of aspartame are essentially inactive biologically (3).

The projected maximum ingestion of aspartame has been estimated to be 22-34 mg/kg/day. The higher dose is calculated as the 99th percentile of projected daily ingestion (2). The 22-34 mg/kg dose range is equivalent to 2.4-3.7 mg/kg/day of methanol, 9.8 mg/kg/day of aspartate, and 12-19 mg/kg/day of phenylalanine (2). The FDA has set the allowable or acceptable daily intake (ADI) of aspartame at 50 mg/kg (1.5 times the 99th percentile of projected daily ingestion) (1,3). In Europe and Canada, the ADI is 40 mg/kg (3). The ADI is defined as an average daily ingestion that is considered harmless, even if continued indefinitely, but does not imply that amounts above this value are harmful (3).

The toxic effects of methanol ingestion are the result of its metabolism to formaldehyde and then to formic acid. Accumulation of this latter product is responsible for the acidosis and ocular toxicity attributable to methanol (1, 2, 3). The dose of methanol estimated to cause significant toxicity is estimated to be 200-500 mg/kg (4). Theoretically, since about 10% of aspartame is methanol, the toxic dose of aspartame, in terms only of methanol, would be about 2000 mg/kg, a dose considered far in excess of any possible ingestion (4). In 12 normal subjects, methanol plasma levels were below the level of detection (0.4 mg/dL) after ingestion of aspartame, 34 mg/kg (3). When abuse doses (100, 150, and 200 mg/kg) of aspartame were administered, statistically significant increases in methanol blood concentrations were measured with peak levels of 1.27, 2.14, and 2.58 mg/dL, respectively (3). Methanol was undetectable at 8 hours after the 100-mg/kg dose but was still present after the higher aspartame doses. Blood and urine formate concentrations were measured after ingestion of 200 mg/kg of aspartame in six healthy subjects (3). No significant increases in blood formate levels were measured, but urinary formate excretion did increase significantly. This indicated that the rate of formate synthesis did not exceed the rate of formate metabolism and excretion (3). No ocular changes or toxicity were observed in the test subjects (3). Moreover, the amount of methanol (approximately 55 mg/L) derived from aspartame-sweetened beverages, containing approximately 555 mg/L of aspartame, is much less than the average methanol content of fruit juices (140 mg/L) (2). Based on the above data, the risk to the fetus from the methanol component of aspartame is nil.

Aspartate is one of two dicarboxylic amino acids (glutamate is the other) that have caused hypothalamic neuronal necrosis in neonatal rodents fed large doses of either the individual amino acids or aspartame (2,5). In neonatal mice, plasma concentrations of aspartate and glutamate must exceed 110 and 75 μmol/dL, respectively, before brain lesions are produced (2). Concerns were raised that this toxicity could occur in humans, especially because glutamate is widespread in the food supply (e.g., monosodium glutamate [MSG]) (6). In addition, some aspartate is transaminated to glutamate, and the neural toxicity of the two amino acids is additive (6). However, brain lesions were not produced in nonhuman primates fed large doses of aspartame, aspartate, or MSG (2).

In normal humans, ingestion of aspartame, 34 mg/kg, or equimolar amounts of aspartate, 13 mg/kg, did not increase plasma aspartate levels (7). Adults heterozygous for PKU also metabolized aspartate normally as evidenced by insignificant increases in plasma aspartate concentrations and unchanged levels of those amino acids that could be derived from aspartate, such as glutamate, asparagine, and glutamine (8). When the dose of aspartame was increased to 50 mg/kg, again no increase in plasma aspartate levels was measured (9). When an abuse dose of aspartame, 100 mg/kg, was administered to adults heterozygous for PKU, plasma aspartate levels increased from a baseline of 0.49 to 0.80 μmol/dL (10). The increase was well within the range of normal postprandial levels. In normal adults given a higher abuse dose of aspartame (200 mg/kg), plasma levels of aspartate plus glutamate were increased from a baseline of 2.7 to 7 μmol/dL, still far below the estimated toxic human plasma level of 100 μmol/dL for aspartate and glutamate (7). Plasma aspartate levels at 2 hours after the 200-mg/kg dose were less than normal postprandial aspartate levels after a meal containing protein (1). Based on these data, subjects heterozygous for PKU metabolize aspartate normally (2). Moreover, neither aspartate nor glutamate is concentrated in the fetus, unlike most other amino acids (1,2,5,11,12). Human placentas perfused in vitro showed a fetal:maternal ratio of 0.13 for aspartic acid and 0.14 for glutamic acid (11). In pregnant monkeys infused with sodium aspartate, 100 mg/kg/h, maternal plasma aspartate levels increased from 0.36 to 80.2 μmol/dL, whereas fetal levels changed from 0.42 to 0.98 μmol/dL (12). Thus, there is no evidence of a risk of fetal aspartate toxicity resulting from maternal ingestion of aspartame, either alone or in combination with glutamate.

High plasma levels of phenylalanine, such as those occurring in PKU, are known to affect the fetus adversely. Phenylalanine, unlike aspartate, is concentrated on the fetal side of the placenta with a fetal:maternal gradient of 1.2:1-1.3:1 (13). The fetus of a mother with PKU may either be heterozygous for PKU (i.e., those who only inherit one autosomal recessive gene and who are nonphenylketonuric) or homozygous for PKU (i.e., those who inherit two autosomal recessive genes and who are phenylketonuric). The incidence of PKU is approximately 1/15,000 persons (5). In contrast, individuals heterozygous for PKU are much more common, with an estimated incidence of 1/50-1/70 persons (8). In the former case, the fetus has virtually no (≤0.3%) phenylalanine hydroxylase activity and is unable to metabolize phenylalanine to tyrosine, thus allowing phenylalanine to accumulate to toxic levels (14). The heterozygous fetus does possess phenylalanine hydroxylase, although only about 10% of normal, with activity of the enzyme detected in the fetal liver as early as 8 weeks' gestation (13,14). Unfortunately, possession of some phenylalanine hydroxylase activity does not reduce the amount of phenylalanine transferred from the mother to the fetus (13). In a study of four families, evidence was found that the heterozygous fetus either did not metabolize the phenylalanine received from the mother or metabolism was minimal (13). This supported previous observations that the degree of mental impairment from exposure to high, continuous maternal levels of phenylalanine is often similar for both the nonphenylketonuric and phenylketonuric fetuses (13). Moreover, the exact mechanism of mental impairment induced by elevated phenylalanine plasma levels has not been determined (4,15,16).

Offspring of women with PKU often are afflicted with mental retardation, microcephaly, congenital heart disease, and low birth weight (16). Pregnancies of these women are also prone to spontaneously abort (16). In one study, maternal phenylalanine plasma levels >120 μmol/dL (classic PKU) were consistently associated with microcephaly, although true mental retardation was observed only when plasma levels exceeded 110 μmol/dL (16). However, research has not excluded the possibility that lower concentrations may be associated with less severe reductions in intelligence (16, 17, 18). For example, in the study cited above, maternal phenylalanine levels <60 μmol/dL (mild hyperphenylalaninemia without urine phenylketones) were associated with normal intelligence in the infants (16). When maternal levels were in the range of 60-100 μmol/dL (atypical PKU), most of the infants also had normal intelligence, but their mean IQ was lower than that of the infants of mothers with mild hyperphenylalaninemia. Others have interpreted these and additional data as indicating a 10.5 point reduction in IQ for each 25.0 μmol/dL rise in maternal phenylalanine plasma concentration (19,20). A recent study, however, examined the nonhyperphenylalaninemic offspring of 12 mothers with untreated hyperphenylalaninemia (21). The results supported the contention that a maternal plasma phenylalanine threshold of 60 μmol/dL existed for an adverse effect on the intelligence of the offspring (21). The investigators, however, were unable to exclude the possibility that nonintellectual dysfunction, such as hyperactivity or attention deficit disorder, may occur at concentrations below the alleged threshold (21). Thus, although this latest study is evidence for a threshold effect, additional studies are needed before the concept of a linear relationship between offspring intelligence and maternal phenylalanine levels can be set aside (19,20,22, 23, 24).

In normal subjects, fasting and postprandial (after a meal containing protein) phenylalanine levels are approximately 6 and 12 μmol/dL, respectively (2,8). When normal adults were administered either a 34- or 50-mg/kg aspartame dose, the mean maximum phenylalanine concentrations were 9-12 and 16 μmol/dL, respectively, with levels returning to baseline 4 hours after ingestion (2,7, 8, 9). Single doses of 100-200 mg/kg, representing abuse ingestions of aspartame, resulted in peak phenylalanine plasma levels ranging from 20 to 49 μmol/dL (2,25). These elevated levels returned to near baseline values within 8 hours. Normal adults were also dosed with an aspartame-sweetened beverage, providing a 10-mg/kg dose of aspartame, at 2-hour intervals for three successive doses (2,26). Plasma levels of phenylalanine rose slightly after each dose, indicating that the phenylalanine load from the previous dose had not been totally eliminated. However, the increases were not statistically significant, and plasma phenylalanine levels never exceeded normal postprandial limits at any time (26). The results of the above studies suggest that even large doses of aspartame do not pose a fetal risk in the normal subject.

Humans heterozygous for the PKU allele metabolize phenylalanine slower than normal persons because of a decreased amount of liver phenylalanine hydroxylase (2). The conversion of phenylalanine to tyrosine is thus impaired, and potentially toxic levels of phenylalanine may accumulate. When adults with this genetic trait were administered aspartame, 34 mg/kg, the mean peak phenylalanine concentration was 15-16 μmol/dL, approximately 36%-45% higher than that measured in normal adults (11 μmol/dL) (2,8,27). To determine the effects of abuse doses, a dose of 100 mg/kg was administered, resulting in a mean peak plasma level of 42 μmol/dL, approximately 100% higher than that observed in normal individuals (20 μmol/dL) (2,10). In both cases, phenylalanine plasma levels were well below presumed toxic levels and returned to baseline values within 8 hours.

A 1986 study examined the effects of a 10-mg/kg dose of aspartame obtained from a commercial product on the basal concentrations of several amino acids, including phenylalanine, in four types of patient: normal, PKU, hyperphenylalaninemic, and PKU carriers (28). One hour after ingestion, mean phenylalanine levels had increased 1.35 μmol/dL (+30%) in normal subjects and 1.35 μmol/dL (+20%) in PKU carriers, decreased 4.58 μmol/dL (−3%) in subjects with PKU, and remained unchanged in hyperphenylalaninemic individuals. The 10-mg/kg dose was selected because it represented, for a 60-kg adult, the dose received from three cans of an aspartame-sweetened soft drink or from approximately 1 quart of aspartame-sweetened Kool-Aid (28).