Potentially Toxic Foods While Breastfeeding: Garlic,Caffeine,Mushrooms,and More

Garlic

Maternal garlic ingestion has a reputation for causing colic in breastfed infants. Two articles tend to refute this claim. In one, 153 mothers who answered a questionnaire were no more likely to report colic in their infants in the previous week if they had ingested garlic than if they had not. In another, mothers who were given either 1.5 g of garlic or placebo in capsules once daily in a blinded manner for 3 days were asked if their infants had exhibited any signs of colic (were fussier, cried more, or had more gas) after capsule ingestion. Four of 20 women who ingested garlic thought their infants had symptoms of colic; however, 4 of 10 women who received placebo thought they had received garlic and reported colic symptoms in their infants.

Coffee, Tea, and Chocolate

With typical intake, the amounts of caffeine in milk are not sufficient to affect the infant. Some experts feel that a maternal intake limit of 300 mg daily (about 2–3 cups coffee) might be a safe level of intake. However, preterm and younger newborn infants metabolize caffeine very slowly and may have serum levels of caffeine similar to their mothers'. Extreme maternal caffeine intake can affect even older breastfed infants as illustrated in the following cases.

Parents of a 24-day-old infant reported restlessness and irritability in their breastfed infant whose mother reported drinking about 0.5 L of yerba mate every other day. Another 6-week-old breastfed infant was reported by a mother to exhibit jitteriness. She claimed to drink four to five cups of coffee and 1–1.5 L of cola daily as well as occasional tea and cocoa. Upon examination, the infant was gaining weight appropriately, but had trembling and increased muscle tone. The infant's symptoms decreased markedly 2 weeks after his mother stopped all caffeine-containing beverages. A third mother noted restlessness and irritability in her breastfed 5-month-old infant while drinking 20 cups of coffee a day.

Two busy physician mothers reported that their breastfed infants had symptoms of stimulation. One reported drinking at least five mugs of coffee, four mugs of tea, and one can of cola daily while breastfeeding two consecutive infants. The first infant slept for only brief periods and woke easily. The second baby was fretful and jumpy and also had poor sleep patterns until his mother stopped her caffeine intake. The other physician, who reportedly drank 1.7–2 L of decaffeinated coffee daily, had premature twins who both seemed to be irritable, although the larger infant was partially supplemented with formula and seemed less affected. When her coffee consumption increased further, the smaller infant suffered convulsive-like episodes. All symptoms resolved 24 hours after coffee consumption ceased. Decaffeinated coffee contains some residual caffeine.

Chocolate contains small amounts of caffeine and much larger amounts of the closely related compound, theobromine. Although low intake of chocolate by a nursing mother is not problematic, extreme amounts can affect the infant. One mother did not drink coffee, but she took about 250 g/day of cocoa and chocolate during pregnancy and lactation. Her infant began breastfeeding 30 minutes postpartum and by 12 hours of life was irritable and jittery, which might partially reflect transplacental passage. These symptoms persisted for days along with inconsolable crying, excessive sucking, and sleep difficulties. Treatment with phenobarbital was ineffective and caffeine was measurable in the infant's blood. The mother tapered her chocolate intake over a 10-day period. The infant's symptoms slowly regressed and neurologic development was normal at 8 months.

Foods and Glucose-6-Phosphate Dehydrogenase Deficiency

The classic food causing hemolysis in glucose-6-phosphate dehydrogenase (G6PD) deficiency is fava beans. Fava beans contain the compounds vicine and convicine. These chemicals are metabolized to divicine and isouramil, which are potent oxidizing agents. In persons with G6PD deficiency, these compounds cause hemolysis by disrupting the red cell wall. Many cases of hemolysis, and subsequent pallor and jaundice, have been reported in breastfed infants after maternal fava bean intake. Most of the cases have been reported from around the Mediterranean and Middle East or in infants whose heritage was from this region. The prevalence of G6PD deficiency is relatively high in this geographic area, where perhaps more susceptible variants occur—at least 14 variants of G6PD deficiency are known. Most reports are of male infants, but some female infants have been affected. Favism through breast milk can be quite severe. One breastfed infant with G6PD deficiency developed renal cortical necrosis after maternal fava bean ingestion and died of renal failure 10 days after maternal fava bean ingestion.

At least one other “food” has caused hemolysis in breastfed infants. Four breastfed infants of three mothers, three boys and one girl, developed severe hemolysis after maternal ingestion of quinine-containing tonic water. The breast milk of one of the mothers was tested and found positive for quinine. All infants had low levels of G6PD and were jaundiced on admission. Cessation of breastfeeding and tonic water plus phototherapy and/or transfusion resolved the jaundice. One of the infants who was severely jaundiced had abnormal brain stem automatized evoked potentials at discharge. At 4 months of age, he had a slight decrease in reactivity and a profound bilateral deafness.

Botulism

Little information is available on the transfer of botulinum toxin into breast milk. Type A botulinum toxin was detected in the blood and stools of a nursing mother after ingesting fermented salmon eggs. No botulinum toxin was detected in the infant's blood or stool on the day the mother was admitted to the hospital and no botulism organisms were detected in the infant's stools. The mother was given two vials of trivalent botulism antitoxin, one intravenously and one intramuscularly. A milk sample obtained 4 hours after administration of the antitoxin had no detectable botulinum toxin nor botulism organisms.

Deadly Mushrooms

Amatoxins are water soluble heat stable polypeptides found in Amanita (most often Amanita phalloides), Galerina, and some Lepiota species. The main toxin from the species A. phalloides is α-amanitin, a cyclic octapeptide. It is a potent inhibitor of RNA polymerases that blocks the production of mRNA and protein synthesis in liver and kidney cells.

Although some articles and textbooks state that amatoxins pass into breast milk, documentation is very poor except for one recent case, which suggests that they might not. A 32-year-old mother shared a meal of foraged mushrooms (Amanita bisporigera), and developed symptoms 15 hours postingestion. She presented to the emergency department (ED) 29 hours postingestion and was found to have markedly elevated liver enzymes. Her 4-month-old-daughter had breastfed 4 hours postingestion. The asymptomatic infant was evaluated 48 hours after breastfeeding and discharged from the ED with no evidence of hepatotoxicity. Amatoxin was undetectable in the milk, although the timing of the milk sample was not stated. Nevertheless, one case is not definitive and the milk of nursing mothers who might have ingested Amanita mushrooms should be considered potentially toxic.

Marine Animal Poisonings

Ciguatera toxin is created in tropical reef-dwelling fish that ingest toxins from blooms of the dinoflagellate Gambierdiscus toxicus. The main toxin appears to be ciguatoxin, although maitotoxin has also been implicated. Ciguatera can present initially with gastrointestinal or neurological symptoms, such as paresthesias of the extremities and around the mouth. Although the fatality rate is low, neurologic effects can last for weeks. A review of 12,890 cases in French Polynesia from 1964 to 1986 found three breastfed infants between 6 and 10 months of age who had been affected after maternal poisoning. In the Bahamas, a case occurred in a 4-month old who was breastfed 1 and 4 hours after maternal ingestion of a kingfish. Symptoms included colic and diarrhea persisting for 48 hours beginning about 10 hours after the first nursing. This was followed by fine maculopapular rash and fussiness that lasted for 2 weeks.

Chelonitoxism is caused by eating sea turtle meat contaminated with chelonitoxins, which are thought to accumulate from the environment without affecting the turtle. Initially, gastrointestinal symptoms occur, followed by neurologic, hepatic, and renal toxicity. One poisoning incident occurred after ingestion of hawksbill turtle, Eretmochelys imbricata, at a family gathering on a remote island in Micronesia. Two adults and four children aged 2–4 years died from the poisoning before reaching the hospital. Two of the four infants who died were breastfed after maternal ingestion, one had received only breast milk and the other had also been given a small amount of turtle meat.

Another cluster of poisonings from eating meat from the same species of sea turtle occurred in western Madagascar and affected 76 victims. Seven lactating women and their seven infants were poisoned. None of the mothers died, but four of the infants died, including one who was exclusively breastfed. Infant symptoms consisted of hypotonia, dyspnea, and coma. Five of the mothers had discontinued nursing their infants because of symptoms, but two of their infants died despite this precaution. Several other breastfed infant deaths have been reported in the literature.

Interestingly, traditional Fijian food taboos prohibit ingestion of potentially toxic marine species, including sea turtle, in pregnant and nursing women. Nursing mothers in this culture have a 60% lower risk of seafood poisoning than the general population. ¹

Many other marine toxins are known, but none have been reported to have been transmitted to a human infant through breast milk. Domoic acid is an extremely toxic chemical from algae that concentrates in shellfish, sardines, and anchovies. In humans, it causes a syndrome called amnesic shellfish poisoning because it adversely affects memory. Domoic acid has been transmitted in the wild to marine mammal pups through their mothers' milk, as well as experimentally in the laboratory through rat milk. The toxin was found in one sea lion's milk 52 days after admission to a rehabilitation facility where ongoing exposure was unlikely, so it may persist in milk for a considerable time after exposure.^2,3

Summary

The passage of food-borne toxins into breast milk is poorly documented, but appears to be possible and quite dangerous with some maternal poisonings. It is safest to assume that all of the most deadly toxins can potentially be transmitted to the infant through breast milk in amounts that can adversely affect the infant. Because the pharmacokinetics of the most potent toxins has not been well characterized in humans, the durations that they persist in milk have not been clearly defined. Mothers thought to be poisoned should probably withhold breastfeeding for several days. More published clinical data on these rare events would be welcome.