Aluminum Contamination of IV Fluids May Cause Brain Damage in Premature Infants

By Laura Ment, MD, and Susan Hockfield, PhD

Preterm birth results in significant developmental disability, and several recent reports have suggested that cognitive outcome may be directly related to gestational age at birth. Infants weighing less than 1500 g at birth now represent 2% of all live births in the United States, and the survival rate of this population of infants is almost 90%. The incidence of major neurodevelopmental handicaps in these infants ranges from 12% to 32%, and these statistics have not changed during the past decade. Thus, identifying and hopefully preventing causes of neurological disability in preterm infants are of utmost importance for pediatricians, neonatologists, and neurologists.

Cerebral development in the human fetus is characterized by sequential periods of neuronal and glial proliferation, migration, and neuronal elaboration of synaptic connection with other cortical and subcortical regions of the brain. By 25 weeks of gestation, the time of birth at which the survival rate was 70% at our institution in 1996, almost all of the developing cortical neurons have been generated, and synaptogenesis is beginning in the developing cortex.

Many neuroscientists believe that the neuronal activity-dependent formation, pruning, and remodeling of these synapses is critical not only for brain growth but also for learning as the infant matures. Prenatal pharmacologic insult such as ethanol and halothane have been shown to decrease synaptic density and impair learning in newborn rats. More recently, lead has been demonstrated to block neuronal transmission at the glutamate receptor, thus suggesting that it also may impair synaptogenesis in the developing brain.

In humans, aluminum toxicity is associated with progressive encephalopathy and has been implicated in Alzheimer’s disease. The mechanisms by which aluminum may alter neuronal function are not yet certain, but two related mechanisms are suggested by a set of recent observations. Neurotransmitters couple to intracellular processes through a variety of second messenger systems. One such second messenger generates diacylglycerol and inositol 1,4,5-triphosphate from inositol 4,5-bisphophate by activation of phospholipase C-1 (PLC-1). This is activated by a subset of glutamate receptors, the class I metabotropic receptors, and a subset of muscarinic acetyl choline receptors, all of which are likely to be important in synaptic plasticity in the developing brain. Aluminum can interfere with this pathway in two ways. Aluminum inhibits PLC-1, and it also reduces the binding of acetyl choline to its receptor. Both of these effects of aluminum could attenuate developmental synaptic plasticity and thus have severe and permanent effects on the establishment of appropriate patterns of connectivity.

Aluminum is known to have neurotoxicity as evidenced in children treated with renal dialysis using aluminum-contaminated dialysis solutions.1 Aluminum toxicity of children on chronic dialysis was undoubtedly aggravated by administration of oral aluminum hydroxide "binders" to reduce levels of serum phosphate. Because commercial intravenous feeding solutions widely used in the United States and United Kingdom are known to be contaminated with aluminum, Bishop and associates investigated possible effects of perinatal exposure to intravenous aluminum on the neurologic development of premature infants.

In a recent study by Bishop et al, 227 premature infants (gestational age < 34 weeks and birth weights < 1850 g) who required intravenous feeding were randomized to receive standard or aluminum-depleted IV feeding solutions. The neurologic development of the 182 survivors who could be tested at 18 months old was assessed using the Bayley Mental Developmental Index (BMDI). Bishop et al estimate that infants receiving the standard IV solutions received 45 mcg/kg/d of aluminum while those receiving the depleted solutions received 4-5 mcg/kg/d.

A total of 90 infants received the standard solution, and 92 received the aluminum depleted solution. There were no significant clinical differences between the two groups except for aluminum intake. When tested at 18 months, there was a small but not significant difference in BMDI between the two groups (95 ± 22 vs 98 ± 20; P = 0.39).

However, in a planned subgroup analysis of infants in whom the duration of intravenous feeding exceeded 10 days, significant differences were seen between the two groups. The BMDI of infants receiving the standard IV feeding solutions was 92 ± 20 compared to 102 ± 17 in those receiving aluminum depleted solutions. This difference was statistically significant (P = 0.02). The former group was more likely to have BMDIs less than 85 (P = 0.03). Bishop et al conclude that there is a loss in BMDIs of about 1% per day in infants receiving the standard, high aluminum IV solution.

This excellent and provocative study indicates that aluminum exposure in the newborn period impairs cognitive outcome in preterm infants at 18 months corrected age. These interesting data suggest that both larger, double-blinded, multicenter, randomized trials and experimental, neonatal animal studies should be performed to confirm the deleterious influence of unmonitored aluminum administration on the developing brain. Even before such a larger study can be mounted, I would not be surprised to find the manufacturers of IV feeding fluids for premature infants on their own begin to reduce the aluminum content of their intravenous products. Since aluminum has no benefit, why not remove it? (Dr. Ment is Professor of Pediatrics and Neurology, and Dr. Hockfield is Professor of Neurology, Yale University School of Medicine.)


1. Bishop NJ, et al. Aluminum neurotoxicity in preterm infants receiving intravenous-feeding solutions. N Engl J Med 1997;336:1557-1561.