Appendix E-2.41 Evidence Portfolio

Large prospective cohort study (Halldorsson et al.) from the Danish National Birth Cohort investigated associations between consumption of artificially sweetened and sugar sweetened soft drinks during pregnancy and subsequent pre-term delivery.

Significant trends in risk of pre-term delivery with increasing consumption of artificially sweetened drinks (carbonated and non-carbonated), but not for sugar-sweetened drinks.

For highest exposure groups (≥ 4 serv/d) vs non-consumption OR = 1.78 (95% CI 1.19-2.66) and 1.29 (95% CI 1.05-1.59) for carbonated and noncarbonated artificially sweetened drinks. Associations with consumption of artificially sweetened carbonated drinks did not differ according to whether delivery was very early (<32 weeks) or only moderately or late pre-term.

The Panel noted the prospective design and large size of study sample were major strengths, and there were no important flaws in the methods used. Panel agreed with the authors who concluded that replication of their findings in another setting was warranted.

Both Halldorsson and Englund-Ögge were well designed and conducted. Noting this, the Panel concluded that even at high levels of exposure to artificially sweetened soft drinks the risk of pre-term delivery is likely to be small, if any. The observed associations could be a consequence of uncontrolled residual confounding.

Large prospective cohort study of Norwegian women (Englund-Ögge et al.) investigated the relationship between consumption of artificially sweetened and sugar-sweetened soft drinks during the first 4-5 months of pregnancy and subsequent pre-term delivery.

No significant trends were found in risk of pre-term delivery with increasing consumption of artificially sweetened drinks or sugar-sweetened drinks. Small elevations of risk were observed with higher consumption of artificially sweetened soft drinks, but after adjustment for covariates, these reached significance only when categories of consumption were aggregated to four levels, then OR = 1.11 (95% CI 1.00-1.24) for the highest category (≥ 1 serving/day) in comparison to non-consumption. This was driven by an increase in spontaneous but not medically induced pre-term delivery. Associations with sugar-sweetened soft drinks tended to be stronger, with an adjusted OR = 1.25 (95% CI 1.08-1.45) for consumption of at least 1 serv/d.

The Panel noted that effects may have been underestimated because of non-differential inaccuracies in the assessment of dietary exposures, but the method was similar to that used by Halldorsson et al. and the same for sugar-sweetened as for artificially sweetened soft drinks.

La Vecchia (2013) performed a meta-analysis of findings from Halldorsson et al. (2010) and Englund-Ögge et al.

The analysis indicated similarly elevated risks of pre-term delivery with higher consumption both of sugar-sweetened and of artificially sweetened drinks.

The lack of specificity in the associations points to possible residual confounding.

Overall, currently available epidemiological data do not suggest that consumption of artificially sweetened soft drinks is a cause of pre-term delivery.

Case-control study (Hardell et. ) in Sweden of 209 patients with brain tumors compared with 425 controls, selected from the Swedish Population Register and matched for sex, age and region of residence. The focus of study was exposure to ionizing radiation/cell phones, but information was also collected on consumption of low-calorie drinks, most of which contained aspartame. 

Non-significant elevations of risk for consumption of low-calorie drinks in relation to brain tumors overall (OR = 1.24, 95% CI 0.72-2.14) and malignant brain tumors specifically (OR = 1.70, 95% CI 0.84-3.44).

The study had a high response rate, but was limited by its relatively small size, the basic assessment of exposure (low-calorie drinks), and the potential for recall bias (because cases knew that they had a brain tumor) all of which could have led to inflation of risk estimates.

Case-control study (Bunin et al.) of 315 US children with medulloblastoma/primitive neuroectodermal tumor diagnosed before the age of 6 y, and 315 control children (selected from the general population by random digit dialing).

In an unadjusted analysis, a significant trend of increasing risk was observed with more frequent consumption of low-calorie carbonated drinks in the pre-conception period. This was attenuated after adjustment for potential confounders, with an adjusted OR = 1.3 (95% CI 0.7-2.5) for ≥2/day versus < 1/month. There were no significant associations with reported frequency of consuming diet soda during midpregnancy.

Assessment of exposure to diet soda required recall after an interval of several years and therefore may not have been reliable. It served only as a proxy for exposure to aspartame (at the time was the most widely used sweetener in soft drinks), and other possible sources of aspartame were not evaluated.

The authors concluded that their results generally did not support an association with aspartame, but the limitations, and also the low statistical power, restrict the conclusions that can be drawn from this study.

Linked set of case-control studies (Gallus et al.) in Italy assessed the association of artificial sweeteners with 9 types of cancer. Patients with incident, histologically confirmed cancers of the oral cavity and pharynx (598), esophagus (304), colon (1225), rectum (728), larynx (460), breast (2569), ovary (1031), prostate (1294) and kidney (767) were compared with 7028 controls admitted to the same hospitals for acute, non neoplastic disorders.

The ORs for consumption of other sweeteners, mainly aspartame, were 0.77 (95% CI 0.39–1.53) for cancers of the oral cavity and pharynx, 0.77 (95% CI 0.34–1.75) for esophageal, 0.90 (95% CI 0.70–1.16) for colon, 0.71 (95% CI 0.50–1.02) for rectal, 1.62 (95% CI 0.84–3.14) for laryngeal, 0.80 (95% CI 0.65–0.97) for breast, 0.75 (95% CI 0.56–1.00) for ovarian, 1.23 (95% CI 0.86–1.76) for prostate and 1.03 (95% CI 0.73–1.46) for kidney cancer. A significant inverse trend in risk for increasing categories of total sweeteners (incl saccharin) was found for breast and ovarian cancer, and a direct one for laryngeal cancer.

Misclassification of exposures will have been non-differential (i.e. similar for cases and controls), in which case the effect will have been to bias risk estimates towards the null. Thus, while the results do not suggest a hazard for the cancers studied, on their own they provide only limited reassurance of safety.

The authors state this study provides no evidence that saccharin or other sweeteners (mainly aspartame) increase the risk of cancer at several common sites in humans.

Case-control update of Gallus et al. in Italy (Bosetti et al.) to test possible associations of artificial sweeteners with 3 types of cancer. Cases were 230 patients with stomach cancer, 326 patients with pancreatic cancer and 454 patients with endometrial cancer. These were compared with 547, 652 and 908 controls, frequency matched by age, sex and study center.

ORs for use of low-calorie sweeteners versus non-use were 0.80 (95% CI, 0.45-1.43) for gastric cancer, 0.62 (95% CI, 0.37-1.04) for pancreatic cancer, and 0.96 (95%CI, 0.67-1.40) for endometrial cancer. Corresponding ORs for saccharin were 0.65 (95% CI, 0.25-1.68), 0.19 (95 % CI, 0.08-0.46), and 0.71 (95% CI, 0.36-1.38), and for other sweeteners were 0.86 (95% CI, 0.45-1.67), 1.16 (95%CI, 0.66-2.04), and 1.07 (95% CI, 0.71-1.61).

The findings do not suggest a hazard, but because exposures to aspartame specifically were not distinguished, and because of possible non-differential misclassification of exposures, on their own they provide only limited evidence of safety.

Authors state that this study adds further evidence on the absence of an adverse effect of low-calorie sweeteners, including aspartame, on risk of common neoplasms.

Case-control study (Andreatta et al.) in Argentina to investigate the relation between consumption of artificial sweeteners and urinary tract tumours (UTT). A study of 197 patients with incident confirmed transitional-cell UTTs along with 397 controls from the same area who had no history of cancer, and had been admitted to hospital with acute non-neoplastic, non-urinary tract diseases.

For long-term use (≥10 y) a positive association was found between use of artificial sweeteners and risk of UTT compared to non-users (OR = 2.18, 95% CI 1.22–3.89, adjusted for age, sex, BMI, social status, and years of tobacco use).

For short-term consumers, no association with UTT was observed.

The Panel noted that ~80% of cases and controls who consumed artificial sweeteners, used saccharin or cyclamate. The study provided little information about possible risks from aspartame.

The authors concluded that use of artificial sweeteners for 10 years or more was positively associated with UTT.

Prospective cohort study by Lim et al. from the NIH-AARP Diet and Health Study included 285,079 men and 188,905 women aged 50-71 y at entry, which was drawn from 8 areas in the US. In this cohort, risk of hematopoietic cancer (1888 cases) and malignant glioma (315 cases) during 5 y follow-up (1995-2000) was examined in relation to daily intake of aspartame assessed at baseline.

During > 5 y follow-up, 1,888 hematopoietic cancers and 315 malignant gliomas.

Higher levels of aspartame intake were not associated with the risk of overall hematopoietic cancer (RR for ≥600 mg/d = 0.98; 95% CI, 0.76-1.27), glioma (RR for ≥400 mg/d = 0.73; 95% CI, 0.46-1.15; P for inverse linear trend = 0.05), or their subtypes in men and women.

The Panel noted that major strengths of this investigation were its prospective longitudinal design, large number of cases, and that exposures were assessed at baseline and therefore unbiased by knowledge of disease outcome. Ascertainment of cancers was reliable. Confounding was unlikely to have been a major problem, although there was no adjustment for socio-economic status (which has shown some relation with brain cancer). Assessment of exposure to aspartame covered aspartame containing drinks as well as addition of aspartame to coffee and tea, but it was limited to one point in time, and usage in the years before the study may have been lower.

The authors concluded that their prospective study suggested that aspartame consumption derived from its main source, aspartame-containing beverages, does not raise the risk of hematopoietic or brain malignancies.

Pilot case-control study (Cabaniols et al.) conducted in France to investigate lifestyle factors and brain cancer risk with 122 incident adult cases of malignant primitive brain tumors (MPBT) and 122 controls with other neurological diagnoses.

There was no association between aspartame consumption during the past five years of at least once per week and risk of MPBT (OR = 1.02, 95% CI 0.57-1.85).

Information on the method of dietary assessment was limited, and there was no attempt to control for potential confounding other than sex and age. In view of this, and the low statistical power of the study (reflected in the CI), the nonpositive finding provides little reassurance of an absence of hazard.

The authors stated that additional large clinical studies are needed to confirm these findings.

Prospective cohort study with Nurses' Health Study and Health Professionals Follow-Up Study cohorts (Schernhammer et al.).

The risk of lymphatic and hematopoietic cancers in relation to consumption of diet soda and aspartame sweeteners added at the table was examined; 77,218 female registered nurses and 47,810 male health professionals were followed for 22 y.

1,324 subjects developed non-Hodgkin lymphoma (NHL), 285 multiple myeloma, and 339 leukemia (mostly myeloid leukemia).

The highest category of aspartame intake (≥143 mg/day) was associated with elevated relative risk of NHL (RR = 1.64, 95% CI 1.17-2.29) and of multiple myeloma (RR = 3.36, 95% CI 1.38-8.19) in men.

There was no consistent trend in risk with increasing exposure, and there were no corresponding elevations in risk in women.

No clear association with leukemia was apparent in either men or women.

Major strengths of this study were its prospective design, the substantial number of cancer cases, the repeated assessment of dietary intake every 4 y,  aspartame intake assessed from time of entry in US diet, and many potential confounders assessed.

The Panel stated that the positive findings can be given little weight, given their limitation to men, the small relative risks observed, and the lack of clear dose-response relationships.

Schernhammer et al speculated that the differential findings for men and women might reflect differences in the activity of alcohol dehydrogenase type 1, which converts methanol (a metabolite of aspartame) to formaldehyde, and is higher in men than in women.

The Panel considered that the results of these epidemiological studies do not suggest an increased risk associated with aspartame consumption for the types of cancer examined.

Randomized controlled trial (Leon et al. 1989) on the effects of aspartame on metabolic outcomes. Participants included 57 women and 51 men randomly assigned to either the aspartame (n = 53, 75 mg/kg/d) or placebo (n = 55) for 24 wks.

No treatment-related hematological changes, alterations in clinical chemistry, urinary abnormalities, or differences in vital signs or body weight.
No differences in blood formate or methanol; urinary Ca2+ or formate; serum folate; or serum lipids.  
No significant changes in amino acid profiles and no evidence of accumulation of phenylalanine or tyrosine.

Controlled diet study (Porikos and Van Italie) with 21 men (24-45 y) given a baseline diet (25-30 % calories from sucrose) alternating with a calorie restricted diet (sweetness replaced with aspartame).

There were small increases in blood urea nitrogen (BUN) in subjects on aspartame containing diet, but renal function remained within normal parameters. Serum triglycerides decreased by 33 % on the aspartame diet.

The Panel noted that the dose of aspartame was not specified and that there were other changes in the diet, including changes in its caloric content.

Overall, the Panel concluded that no significant adverse effects were observed following repeated administration of aspartame for different durations and at different dose levels of aspartame in healthy subjects.

Cross-over study (Kruesi et al.) investigated the effect of sugar and aspartame consumption on behavior in 30 preschool boys (2 - 6 y). Double blind cross-over challenge with aspartame (30 mg/kg bw), sucrose (1.75 g/kg bw), saccharin (amount not specified) and glucose (1.75 g/kg bw).

There was no significant difference in scores of aggression or observers’’ ratings of behavior between the sugar responsive and age-matched control boys following any of the four treatments.

Randomized cross-over trial (Wolraich et al.) with 25 normal preschool children (3 - 5 y) and 23 primary school-age children (6 - 10 y). Over three consecutive 3-wk periods, double blind randomized distribution to a diet high in aspartame (32-38 mg/kg/d), sucrose (4500-5600 mg/kg/d), or saccharin (10-12 mg/kg/d).

For the children described as sugar-sensitive, there were no significant differences among the 3 diets in any of the 39 behavioral and cognitive variables. For the preschool children significant differences were measured in 4 of the 31 measures (Parents’’ ratings of cognition, grooved pegboard, dominant hand, non-dominant hand); however, no consistent pattern in behavioral and cognitive differences was observed amongst the 3 diets.

The authors concluded that ‘even when intake exceeded typical dietary levels, neither dietary sucrose nor aspartame affects children’s behavior or cognitive function’.

Randomized crossover trial (Shaywitz et al.) of the effect of large doses of aspartame on behavior, cognitive function and monoamine metabolism in 15 children with attention deficit disorder (11 boys and 4 girls, 5-13 y). The trial consisted of two 2-week periods that were identical except for the administration of either aspartame (34 mg/kg/d) or placebo (microcrystalline cellulose).

No significant effect of aspartame on cognitive, attentive or behavioral testing. The biochemical and hematological parameters were not altered by aspartame except that plasma phenylalanine levels increased by approximately 40 % two hours following aspartame administration (within in normal postprandial range).

The authors state that the findings indicate that aspartame at greater than 10 times usual consumption had no effect on the cognitive and behavioral status of children with attention deficit disorder.  In addition, aspartame did not appear to affect urinary excretion rates of monoamines and metabolites.

Non-randomized, controlled study of Roshon and Hagen (1989) examined the effect of sucrose consumption on the behavior of 12 preschool children (6 boys and 6 girls, 3-5 y).

No significant difference in locomotion, task orientation and learning in participants exposed to either sucrose or placebo (9 mg aspartame/kg).

The Panel noted that no effects of aspartame on behavior and cognition were observed in children in these studies.

Randomized, crossover trial (Lapierre et al.) with 10 healthy adults (6 men, 4 women, 21––36 y) who received a single dose of aspartame (15 mg/kg) or placebo capsules.

No significant differences between aspartame and placebo were found in measures of sedation, hunger, headache, reaction-time, cognition or memory during the study. Plasma phenylalanine levels rose within thirty minutes of administration of aspartame.

Randomized crossover trial (Ryan-Harshman et al.) with healthy males age 20 - 35 y (n = 13/group) given phenylalanine capsules (0.8, 2.5, 5 and 10 g) or aspartame (5 or 10 g) as a single dose to investigate neurobehavioral effects on energy and macronutrient selection and on subjective feelings of hunger, mood and arousal.

Neither phenylalanine nor aspartame altered mean energy intakes or macronutrient selection nor caused any behavioral effects.

Non-randomized, controlled crossover study (Pivonka and Grunewald 1990) examined the effect on mood and well-being in 120 women (18 - 30 y) receiving water, aspartame-sweetened or sugar-sweetened beverages.

Mood tests employed were the Stanford Sleepiness Scale (SSS), the Visual Analogue Mood Scale (VAMS), and the Profile of Mood States (POMS). Changes in mood were similar following consumption of water or the aspartame-sweetened beverage. However, the ingestion of the sugar-sweetened beverage was followed by increased sleepiness during the last half of the one-hour observation period (p less than .002).

Double-blind study (Stokes et al.), 12 healthy pilots (4 females and 8 males) were given placebo, aspartame (50 mg/kg bw) or ethanol (positive control, dose not reported but estimated to raise plasma alcohol to 0.1 %). Each subject performed the SPARTANS cognitive test battery of aviation-relevant information-processing tasks on 5 sessions after a single treatment.

No detectable performance decrements were associated with the exposure to aspartame, but decrements in psychomotor and spatial abilities were detected following ethanol administration.

Follow up study (Stokes et al.) was undertaken in 12 subjects (college students, sex not reported) in order to examine the effects of double-blind repeated dosing of aspartame on performance in aviation-relevant cognitive tasks.

The subjects received placebo capsules or aspartame capsules (50 mg/kg bw/day) for 9 days, or an acute dose of ethanol to achieve 0.1 % blood ethanol levels.

Forty-seven task variables were measured using the SPARTANS 2.0 cognitive test battery and no significantly impaired performance on flight-relevant cognitive tasks was observed.

Study by Walton et al. was designed to test whether subjects with mood disorders were sensitive to adverse effects caused by aspartame; 40 adult patients with unipolar depression and a similar number of adult subjects without a psychiatric history were recruited. The participants were given aspartame (30 mg/kg bw/day) or placebo (sucrose) in capsules for a period of 7 days with two 3-day washout periods using a double-blind cross-over study design.

Despite the small number of subjects, there was a significant difference in the number and severity of self-scored symptoms between aspartame and placebo in the patient group while there was no difference in the non-depressed volunteer group.

Randomized double-blind placebo-controlled 3-way crossover, 48 adults (24 men, 24 women, 18––34 y) were exposed after an initial one-month aspartame-free period, to aspartame, sucrose or placebo administered for 20 days each (Spiers et al.). Twenty-four participants were given a high dose of aspartame (45 mg/kg/d) and the remaining received a low dose of aspartame (15 mg/kg/d). The dose of sucrose was 90 g/day for all subjects. Administration of aspartame or placebo (microcrystalline cellulose) was in the form of capsules and for sucrose, a beverage.

Plasma phenylalanine levels increased dose-dependently with aspartame consumption from 56 μM (placebo) to 79 μM (high dose), but no neuropsychologic, neurophysiologic and behavioral effects linked to aspartame consumption were observed.

The Panel noted the limited number of participants, the short duration and the inconsistency of the reporting of the results in all these adult human studies. These limitations apply to both positive and negative studies.
Overall, the Panel concluded that there was no evidence that aspartame affects behavior or cognitive function in children or adults.

Double-blind study in 8 girls and 2 boys (5.1 - 14.6 y) diagnosed with generalized absence seizures (petit mal seizures) to investigate whether aspartame exacerbates occurrence of seizures (Camfield et al.). Following 1-hr baseline recordings of the number and length of spike-wave bursts (determined using an ambulatory cassette EEG recorder), the children were given 250 ml orange juice sweetened with either aspartame (40 mg/kg) or sucrose (1 g sucrose for every 25 mg aspartame to achieve similar sweetness) (assigned randomly) and the EEG recordings were continued for 6 h. Each child was tested once with each substance.

Following the consumption of aspartame but not of sucrose, the total duration of spike-wave discharge per hr was significantly increased and aspartame appeared to exacerbate the amount of EEG spike waves in children with absence seizures.

The Panel noted that the combination of the two parameters (number and length of spike-wave bursts) into a single measure was not adequately explained, and lack of control of food and drink intake before and after dosing may have affected the results. The Panel further noted that aspartame was given in a single dose at the ADI.

Randomized double-blind placebo-controlled, crossover study (Shaywitz et al.) aspartame (34 mg/kg) was administered to epileptic children (5 boys, 5 girls, 5-13 y) for 2 wks to investigate the induction of seizures following aspartame consumption.

Nine children completed the study and it was reported that there was no difference in the occurrence of seizures between aspartame and placebo exposure. The plasma levels of phenylalanine increased from 60 μM to 82 μM by one hour post aspartame administration.

Randomized double-blind placebo-controlled crossover study by Rowan et al. In this trial, subjects (sixteen adults and two children) who claimed to have experienced epileptic seizures reportedly due to aspartame were given capsules either containing microcrystalline cellulose (placebo) or aspartame (total dose of 50 mg/kg bw). This dose was divided into three portions and administered in the morning at two-hour intervals

The authors reported no seizures or other adverse effects from aspartame ingestion. Mean plasma phenylalanine levels increased from 52 μM (after placebo) to 84 μM two hours after the first two doses aspartame.

Overall the Panel concluded that the available data do not provide evidence for a relationship between aspartame consumption and seizures.

Schiffman et al. reported a randomized double-blind crossover trial with aspartame on 40 subjects with a history of headache and related neurologic symptoms within 24h of aspartame consumption. The subjects (12 males and 28 females, 19––69 y) were given aspartame (30 mg/kg bw) or placebo (microcrystalline cellulose) in capsules; the dose was divided into 3 doses administered in the morning at 2 h intervals.

The incidence rate of headache after consumption of aspartame (35%) was not significantly different from that after placebo (45%).

Randomized cross-over trial by Koehler and Glaros comparing the effect of aspartame to matched placebo on frequency and intensity of migraine headache. The subjects (2 males, 8 females; 18 - 47 y) who had medical diagnosis of migraine, consumed aspartame (1200 mg/person) or placebo (microcrystalline cellulose) in capsules and during two 4 wk phases.

Statistical analysis indicated a significant increase in the frequency of migraine headaches from the placebo to the aspartame treatment (mean number of migraines per subject: 1.72 (baseline phase), 1.55 (placebo phase), and 3.55 (aspartame phase)). No differences were reported in the intensity or duration of migraine headaches.

The high drop-out rate, from 25 to 11 participants in this study was not due to increased frequency or intensity of migraines.

The Panel noted that the high inter individual variability in the response of the remaining volunteers makes interpretation unreliable.

In the study by Lipton et al. 171 patients at a headache unit completed a survey in which alcohol, aspartame, or carbohydrates intake were felt to be triggers of their headaches.

Study showed that 8% reported aspartame as a trigger of headaches compared to 2.3% for carbohydrates, and to about 50% for alcohol.

The Panel considered that having only listed possible triggers of headaches was a major limitation of this study.

Van den Eeden et al. conducted a double-blind randomized cross-over trial with subjects self-diagnosed as sensitive to aspartame. Of the 32 subjects recruited and randomized to receive aspartame (in capsules given 3X/d to give a daily dose of 30 mg/kg bw/day for 7 d) and placebo (microcrystalline cellulose in capsules given 3X/d), only 18 participants completed the full study.

Participants reported headaches on 33 % of days during aspartame intake, compared with 24 % on placebo treatment (p = 0.04). However, no significant difference in the length or intensity of headaches or in the occurrence of side effects associated with the headaches was observed between treatments.

The authors concluded that a small subset of the population may be susceptible to headaches induced by aspartame.

The possible effect of aspartame on headaches has been investigated in various studies, which reported conflicting results, ranging from no effect to the suggestion that a small subset of the population may be susceptible to aspartame-induced headaches. The number of existing studies was small, and several had high participant drop-out rates, both under placebo and aspartame treatment. Overall, the Panel noted that because of the limitations of the studies it is not possible to conclude on a relationship between aspartame consumption and headaches.

The Panel is aware that a number of studies have focused on the effects of aspartame on appetite, hunger and food intake.
The Panel considered that these studies of the effect of aspartame (or other low calorie sweeteners) on eating behavior were not relevant for the assessment of the safety of aspartame and that risk benefit assessment of aspartame are not within the term of reference and the remit of the Panel.

Study by Szucs et al. on aspartame effects on mast cells in vitro.

Aspartame did not affect IgE-mediated histamine release from mast cells in vitro.
Mast cells cultured in the presence of aspartame for up to 9 days showed enhanced proliferation and decreased responsiveness to releasing stimuli.

The authors concluded that the effect of aspartame on proliferation of cells in culture could be ascribed to a nonspecific enhancing effect of its constituent amino acids. Aspartame did not stimulate mast cell or basophil in vivo as assessed by skin testing.

Kulczycki reported a cases of aspartame induced urticaria.

Reported a case of aspartame induced urticaria confirmed by double blind challenge in a 23 year old woman with no history of allergic disease. A second case in a 42 year old woman was also reported.

Garriga et al. (1991) attempted to identify subjects with hypersensitivity reactions to aspartame with blinded challenge procedures. A total of 61 self-referrals and physician referrals were screened, with 20 referrals evaluated in the clinic. Twelve patients underwent single- and double-blind challenge with up to 2000 mg of aspartame.

No subject with a clearly reproducible adverse reaction to aspartame was identified.

The authors concluded that subjects who believed themselves to be allergic to aspartame did not have reproducible reactions.

Geha et al. conducted a multi-centre placebo-controlled clinical study to evaluate individuals who had experienced urticaria and/or angioedema associated with ingestion of food containing aspartame. In a double-blind crossover study, 21 recruited subjects with a self-reported history of hypersensitivity to aspartame were exposed to aspartame and placebo. Conversion products of aspartame, aspartyl-phenylalanine diketopiperazine and beta-aspartame, were also included in the study. Patients received, on different days, increasing doses of aspartame (50, 300, 600 mg) and placebo.

Four urticaria reactions were observed, two followed aspartame ingestion and two followed
placebo ingestion.

The authors concluded that aspartame and its conversion products were no more likely than placebo to cause allergic symptoms in subjects with a history consistent with hypersensitivity to aspartame.

Butchko et al. reviewed all published papers from 1980 onwards reporting allergic-type reactions and attributed to aspartame exposure.

In an evaluation of consumer complaints related to aspartame by the Centers for Disease Control and Prevention (CDC, 1984) approximately 15 % of the anecdotal complaints were assigned to allergic-dermatologic reactions attributed to aspartame ingestion, such as rashes, sore throat/mouth, swelling and itching.
Cases of urticaria and granulomatous panniculitis thought to be related to aspartame were reported.

The Panel noted that the studies available were performed on a limited number of participants.
Overall, taking into account the limited data currently available, the Panel considered that the weight of evidence does not suggest that aspartame is associated with allergic-type reactions in experimental models or in humans. However, the Panel cannot exclude the possibility that in rare instances individuals could be susceptible to allergic reactions following aspartame ingestion.

Prospective cohort study by Maslova et al. explored how intake of artificially sweetened beverages during pregnancy related to asthma and allergic rhinitis in children at 18 months and 7 years of age. Analysis was based on 60,466 pregnant women who enrolled in the prospective longitudinal Danish National Birth Cohort between 1996 and 2002.

In comparison with no consumption of artificially sweetened non-carbonated soft drinks during pregnancy, consumption of at least one serving per day was associated with an increased risk of asthma by three of the four case definitions (odds ratios up to 1.23, 95% CI 1.13-1.33 for asthma at 18 months), but there was no consistent exposure-response relationship across lower frequencies of consumption.
In a corresponding analysis for artificially sweetened carbonated soft drinks, elevated odds ratios were observed for all four case definitions with the highest odds ratios of 1.30, 95% CI 1.01 –– 1.66 for asthma at 7 years of age identified through the Danish National Patient Registry, but again without clear exposure-response relationships. Allergic rhinitis was non-significantly associated with daily consumption of artificially sweetened carbonated soft drinks (OR 1.31, 95% CI 0.98-1.74); no association was observed with daily consumption of artificially sweetened non-carbonated drinks (odds ratios of 1.03, 95% CI 0.86-1.24).

Because in epidemiological terms, the elevations of risk were only small and inconsistent, the findings from this study can only be considered weakly suggestive of hazard i.e. an association between the consumption of artificially sweetened beverages during pregnancy and the diagnosis of asthma or allergic rhinitis in children. Before a final conclusion can be reached with regard to aspartame, the findings need to be explored further with more detailed assessment of exposure to specific artificial sweeteners.

The Panel considered that adverse effects on reproduction and development reported for aspartame in animal studies could be attributed to the metabolite phenylalanine. They undertook a formal Mode of Action (MoA) analysis of the putative role of phenylalanine in developmental toxicity (as seen in animal studies).

Risk characterization was based on comparison of plasma phenylalanine levels following aspartame administration with plasma phenylalanine levels associated with developmental effects in children born from mothers with PKU.
Data on the concentrations of phenylalanine in plasma after different doses of aspartame were extracted from various studies, mainly unpublished studies submitted in response to EFSA’s call for data.

Current clinical practice guidelines recommend PKU patients restrict dietary intake of phenylalanine to keep plasma levels below 360  μM. The Panel noted that intakes of aspartame as a food additive could occur at the same time as other dietary phenylalanine sources. Therefore, they considered the threshold utilized for comparisons should be lowered to allow for simultaneous intake of the food additive with meals.
The highest mean dietary phenylalanine exposure per meal is 34.2 mg/kg bw and this corresponds to a phenylalanine plasma concentration of 120 μM. So plasma phenylalanine from the diet (120 μM) was subtracted from 360 μM to determine the maximum safe plasma concentration of phenylalanine that can be derived from aspartame (240 μM).

The Panel considered that given the conservative assumptions, realistic dietary intake of aspartame and the confidence intervals provided by the modeling, the peak plasma phenylalanine levels would not exceed the clinical target threshold when a normal individual consumed aspartame at levels below the current ADI of 40 mg/kg bw/day.

The Panel therefore concluded that there would not be a risk of adverse effects on pregnancy in the general population including heterozygous individuals at the current ADI.

Overall Conclusions
The Panel concluded that chronic toxicity and reproductive and developmental toxicity were the critical endpoints in the animal database. The Panel considered that the evaluation of long-term effects of aspartame should continue to be based on the animal data. Based on a Mode of Action analysis and the weight-of-evidence, the Panel considered that the reproductive and developmental toxicity in animals was due to phenylalanine released from aspartame and concluded that the basis for evaluation of the reproductive and developmental endpoint should be the available data in humans.

Conservative estimates of exposure to aspartame made by the Panel for the general population were ≤36 mg/kg bw/day at the 95^th percentile.

Due to the conservatism of both the exposure assessment and the aspartame dose-phenylalanine concentration response modeling, the Panel considered that it was highly unlikely that any individual in the normal and PKU heterozygous population would have plasma levels of phenylalanine above 240 uM following oral aspartame exposure up to the ADI of 40mg/kg bw/day. The Panel further considered that even in combination with diet, these aspartame intakes would not lead to peak plasma phenylalanine concentrations above 360 uM, the current clinical guideline for prevention of adverse effects on the fetuses of PKU mothers.

The Panel concluded from the present assessment of aspartame that there were no safety concerns at the current ADI of 40 mg/kg bw/day. Therefore, there was no reason to revise the ADI for aspartame.

The Panel emphasized that its evaluation of phenylalanine plasma levels from a dose of aspartame at the ensuing ADI is not applicable to PKU patients. These individuals require total control of dietary phenylalanine intake to manage the risk from elevated phenylalanine plasma levels.