Journal Article
Elizabeth L Prado,
Department of Nutrition, University of California at Davis, Davis, CA, USA
SUMMIT Institute of Development, Mataram, Nusa Tenggara Barat, Indonesia
Correspondence: EL Prado, Program in International and Community Nutrition, University of California at Davis, 3253 Meyer Hall, One Shields Ave, Davis, CA 95616, USA. E-mail: . Phone: +1-530-752-1992. Fax: +1-530-752-3406.
Search for other works by this author on:
Kathryn G DeweyDepartment of Nutrition, University of California at Davis, Davis, CA, USA
SUMMIT Institute of Development, Mataram, Nusa Tenggara Barat, Indonesia
Search for other works by this author on:
-
- Split View
-
- Article contents
- Figures & tables
- Video
- Audio
- Supplementary Data
-
- More
Close
Navbar Search Filter Microsite Search Term Search
Abstract
Presented here is an overview of the pathway from early nutrient deficiency to long-term brain function, cognition, and productivity, focusing on research from low- and middle-income countries. Animal models have demonstrated the importance of adequate nutrition for the neurodevelopmental processes that occur rapidly during pregnancy and infancy, such as neuron proliferation and myelination. However, several factors influence whether nutrient deficiencies during this period cause permanent cognitive deficits in human populations, including the child's interaction with the environment, the timing and degree of nutrient deficiency, and the possibility of recovery. These factors should be taken into account in the design and interpretation of future research. Certain types of nutritional deficiency clearly impair brain development, including severe acute malnutrition, chronic undernutrition, iron deficiency, and iodine deficiency. While strategies such as salt iodization and micronutrient powders have been shown to improve these conditions, direct evidence of their impact on brain development is scarce. Other strategies also require further research, including supplementation with iron and other micronutrients, essential fatty acids, and fortified food supplements during pregnancy and infancy.
Introduction
Adequate nutrition is necessary for normal brain development. Nutrition is especially important during pregnancy and infancy, which are crucial periods for the formation of the brain, laying the foundation for the development of cognitive, motor, and socio-emotional skills throughout childhood and adulthood. Thus, nutritional deficiencies during pregnancy and infancy are likely to affect cognition, behavior, and productivity throughout the school years and adulthood. Focusing on this early period for the prevention of nutrient deficiencies may have long-term and widespread benefits for individuals and societies.
This article presents an overview of the pathway from early nutritional deprivation to long-term brain function, cognition, behavior, and productivity. Although nutrition is important for brain function throughout the lifespan, this article focuses on nutrition during pregnancy and the first few years after birth, which is the period of most rapid brain development. Presented first are the biological mechanisms through which nutrient deficiencies in pregnancy and infancy may affect brain development. Most of this evidence at the cellular and molecular level is from animal studies. Although these animal models have demonstrated the importance of adequate nutrition for the developing brain, many factors influence whether undernutrition during pregnancy and infancy leads to permanent cognitive deficits in human populations. The second part of this article discusses four of those factors: 1] the amount and quality of stimulation the child receives from the environment; 2] the timing of nutrient deprivation; 3] the degree of nutrient deficiency; and 4] the possibility of recovery. Finally, a brief review of human studies is presented, focusing on research from low- and middle-income countries, where multiple nutrient deficiencies are prevalent among pregnant women and children.1 Also addressed in this review are the long-term consequences of undernutrition in early life, randomized trials of food and protein/energy supplementation, and studies of breastfeeding practices, essential fatty acids, and certain specific micronutrients, in addition to implications for policy, programs, and future research.
Role of Nutrients in Brain Development
Approximately 22 days after conception, the neural plate begins to fold inward, forming the neural tube, which eventually becomes the brain and spinal cord.2 Adequate nutrition is necessary from the beginning, with the formation of the neural plate and neural tube affected by nutrients such as folic acid, copper, and vitamin A. Seven weeks after conception, cell division begins within the neural tube, creating nerve cells [neurons] and glial cells [cells that support neurons]. After a neuron is created, it migrates to its place in the brain, where it then grows axons and dendrites projecting out from its cell body. These branching projections make connections with other cells, called synapses, through which nerve signals travel from one cell to another. These neurodevelopmental processes begin during gestation and continue throughout infancy [see Table 1]. Groups of neurons form pathways, which are refined through the programmed elimination of cells and connections. About half of all the cells that are produced in the brain are subsequently eliminated throughout childhood and adolescence. Synapses are also overproduced and then selectively eliminated. Some of this refining of neural pathways depends on the child's experience, or in other words, input from the child's environment. Cells and connections that are activated are retained and strengthened while those that are not used are eliminated. This is thought to be one of the primary mechanisms of brain plasticity, allowing the brain to organize itself to adapt to the environment and reorganize itself to recover from injury during development.2
Table 1
Evidence for the role of selected nutrients and experience in five key neurodevelopmental processes
| Definition and timing | Neuron proliferation is the creation of new cells through cell division. This begins in week 7 of gestation and continues to at least 4.5 months postpartum.2 Neuron proliferation is mostly completed at birth, but neurons can be created in adulthood.3 | Axons and dendrites are branching projections that grow out from cell bodies to make connections with other cells. This process begins during gestation and continues through at least 2 years after birth. In some brain areas, axons reach their final destinations at 15 weeks gestation, in others at 32 weeks gestation. Dendrite growth begins at 15 weeks gestation and continues through the second year after birth in some brain areas.2 | Synapses are connections between axons, dendrites, and cell bodies. Synapse formation begins during gestation [around week 23] and continues throughout the lifespan. Synaptic density reaches a peak at different times in different brain areas [for example, in the visual cortex between 4 and 12 months postpartum, and in the prefrontal cortex after 15 months postpartum].3 The decrease in synaptic density that follows this peak in each area reflects synaptic pruning. Synapse overproduction is completed in the second year after birth, while synaptic pruning begins in the first year after birth and continues through adolescence.2 | Myelin is white, fatty matter that covers axons and accelerates the speed of nerve impulses traveling from one cell to another. Myelination begins as early as 12–14 weeks of gestation in the spinal cord and continues until adulthood. The most significant period of myelination occurs from mid-gestation to age 2 years.4 Before birth, myelination occurs in brain areas involved in orientation and balance. After birth, the rate of myelination of areas involved in vision and hearing reaches a peak before myelination of areas underlying language, coinciding with the emergence of these abilities.2 | Apoptosis is programmed cell death. Of all the cells that are produced in the brain, about half die through a variety of mechanisms. One of these mechanisms is programmed cell death, which is regulated primarily by neurotrophic factors, such as BDNF and IGF-1. When levels of neurotrophic factors are below a certain threshold, molecules within the cell trigger degeneration. Neuron apoptosis coincides with the period of synaptogenesis, beginning during gestation and continuing through adolescence.5 Small head size or brain volume may be caused by decreased neuron proliferation or increased apoptosis. |
| Protein-energy malnutrition | Human autopsy studies and magnetic resonance imaging studies have shown that infants with IUGR had fewer brain cells and cerebral cortical grey matter volume than normal-birth-weight infants.6,7 | A human autopsy study showed that 3–4-month-old infants with moderate malnutrition [low weight for age] had decreased dendritic span and arborization [complexity of branching projections] compared to well-nourished infants.9 Rodent models have found similar effects of early postnatal undernutrition on dendrite growth.10 | Both prenatal and postnatal undernutrition in rodents results in fewer synapses as well as synaptic structural changes.a11,12 | Adults who had been exposed to famine in utero in Holland during World War II showed increased white matter hyperintensities, shown by MRI.13 Reduced myelination has been found in animal models of IUGR6 and maternal nutrient restriction without resulting in IUGR.14 IUGR decreases IGF-1 levels and IGF-1-binding protein expression, which influence myelin production.6 | BDNF and IGF-1 levels decreased and cell death increased in the offspring of baboon mothers fed a nutrient-restricted diet during pregnancy [without IUGR].14 |
| Human autopsy studies have also shown that infants with severe acute malnutrition have fewer brain cells than well-nourished infants.8 IUGR in animals results in similar effects.6 | Other animal models have also shown that IUGR decreases IGF-1 and IGF-1-binding protein expression.6 | ||||
| Fatty acids | Neurogenesis requires the synthesis of large amounts of membrane phospholipid from fatty acids. Reduced neuron proliferation has been shown in animals with gestational DHA deficiency.15 | Arachidonic acid and docosahexaenoic acid [DHA] in membranes at synaptic sites play a role in the maturation of synapses and in neurotransmission.16 | Fatty acids are structural components of myelin. Both prenatal and postnatal fatty acid deficiency in rodents reduces the amount and alters the composition of myelin.17,18 | ||
| Iron | Iron is required for the enzyme ribonucleotide reductase that regulates central nervous system cell division.6 While gestational and neonatal iron deficiency in rodents does not affect overall brain size, a decrease in the size of the hippocampus [a subcortical structure that underlies learning and memory] has been shown.19 | Gestational and neonatal iron deficiency in rodents results in truncated dendritic branching in the hippocampus, which persists into adulthood despite iron repletion.20 | Gestational and early postnatal iron deficiency in rodents results in decreased synaptic maturity and efficacy in the hippocampus, which persists despite iron repletion.21 | Iron plays a role in myelin synthesis. In animal models, even marginal iron deficiency during prenatal and early postnatal development decreases myelin synthesis and alters myelin composition, which is not corrected with iron repletion.24 | |
| In adult rodents, iron deficiency decreases the number of dopamine D2 receptors and the density of dopamine transporter in the striatum and nucleus accumbens.22 | |||||
| In both animal models and cell culture experiments, dopamine and norepinephrine metabolism are altered by iron deficiency.23 | |||||
| Iodine and thyroid hormones | Some fetuses aborted in months 6 and 8 of gestation in an iodine-deficient area of China had lower brain weight than fetuses in an iodine-sufficient area, while some showed increased cell density. Gestational iodine deficiency in sheep and marmosets resulted in reduced brain weight and cell number, which was not corrected with iodine repletion. No effect on brain weight or cell number was found in rodents with gestational iodine deficiency, but cell migration was impaired.25 | Gestational iodine deficiency results in reduced dendritic branching in the cerebral cortex in rodents26 and in the cerebellum in sheep and marmosets.25 | Gestational iodine deficiency in sheep resulted in decreased synaptic density, which was not corrected with iodine repletion.25 | Gestational and early postnatal hypothyroidism in rodents decreases the number and density of synapses in the cerebellum, and alters neurotransmitter levels.27 | No myelination was detected in the cerebral cortex of fetuses aborted at month 8 of gestation in an iodine-deficient area of China.28 Gestational iodine deficiency in sheep and rodents reduces myelination.25 |
| Early postnatal hypothyroidism in rodents results in decreased dendritic branching in the visual and auditory cortex and cerebellum.27 | Gestational and early postnatal hypothyroidism in rodents leads to reduced myelination.27 | ||||
| Zinc | Zinc is necessary for cell division due to its role in DNA synthesis. Gestational zinc deficiency in rodents results in decreased number of cells, as reflected by total brain DNA29 and reduced regional brain mass in the cerebellum, limbic system, and cerebral cortex.6 | Gestational zinc deficiency in rodents results in reduced dendritic arborization.6 | Zinc released into synapses in the hippocampus and cerebral cortex modulates synapse function. Specifically, zinc modulates postsynaptic NMDA receptors for glutamate and inhibits GABAB receptor activation.30 | In rodent pups, zinc deficiency decreased expression of IGF-1 and growth hormone receptor genes.31 | |
| Choline | Choline is essential for stem cell proliferation and is involved in transmembrane signaling during neurogenesis.6 In rodents, gestational choline supplementation stimulates cell division.32 | The neurotransmitter acetylcholine is synthesized from choline. Gestational choline deficiency in rodents has long-term effects on cholinergic neurotransmission despite repletion.32 | Gestational choline deficiency increases the rate of apoptosis in the hippocampus in rodents.32 | ||
| B-vitamins | Before neuron proliferation begins, during weeks 2−4 of gestation, the neural tube forms, which is comprised of progenitor [stem] cells that give rise to neurons and glial cells [cells that support neurons].2 Maternal deficiency in folic acid and vitamin B12 is associated with neural tube defects, such as anencephaly and spina bifida.33 | Gestational and early postnatal vitamin B6 deficiency in rodents results in reduced dendritic branching in the neocortex and cerebellum.34,35 | Gestational and early postnatal vitamin B6 deficiency in rodents results in decreased synaptic density in the neocortex,35 reduced synaptic efficiency, particularly in NMDA receptors,36 and lowered dopamine levels and dopamine D2 receptor binding in the striatum.37 | Gestational and early postnatal vitamin B6 deficiency in rodents results in reduced myelination.38 | |
| Experience | Rodents raised in enriched environments [large enclosures with objects that allow visual and tactile stimulation] show greater brain weight and cortical thickness than rodents raised in impoverished environments [standard lab cages].39 | A human autopsy study showed that individuals with higher levels of education had more dendritic branching than those with lower education in Wernicke's area, a brain area underlying language processing.40 Rodents raised in enriched environments [filled with toys and other rodents] have more dendritic spines than those raised in less complex environments.41 | Rodents raised in enriched environments [large enclosures with objects that allow visual and tactile stimulation] show more synapses per neuron in visual and motor cortices than rodents raised in impoverished environments [standard lab cages].39 | Children raised in Romanian orphanages and then adopted into US families, thus having experienced a degree of early socioemotional deprivation, showed structural changes in white matter tracts compared to control children who had not spent any time in an orphanage.42 | |
| Practicing the piano in childhood correlated with myelination in areas underlying finger movements, as measured by fractional anistropy.43 | |||||
| An enriched rearing environment affects myelination of the corpus callosum in rodents and monkeys.44,45 |
| Definition and timing | Neuron proliferation is the creation of new cells through cell division. This begins in week 7 of gestation and continues to at least 4.5 months postpartum.2 Neuron proliferation is mostly completed at birth, but neurons can be created in adulthood.3 | Axons and dendrites are branching projections that grow out from cell bodies to make connections with other cells. This process begins during gestation and continues through at least 2 years after birth. In some brain areas, axons reach their final destinations at 15 weeks gestation, in others at 32 weeks gestation. Dendrite growth begins at 15 weeks gestation and continues through the second year after birth in some brain areas.2 | Synapses are connections between axons, dendrites, and cell bodies. Synapse formation begins during gestation [around week 23] and continues throughout the lifespan. Synaptic density reaches a peak at different times in different brain areas [for example, in the visual cortex between 4 and 12 months postpartum, and in the prefrontal cortex after 15 months postpartum].3 The decrease in synaptic density that follows this peak in each area reflects synaptic pruning. Synapse overproduction is completed in the second year after birth, while synaptic pruning begins in the first year after birth and continues through adolescence.2 | Myelin is white, fatty matter that covers axons and accelerates the speed of nerve impulses traveling from one cell to another. Myelination begins as early as 12–14 weeks of gestation in the spinal cord and continues until adulthood. The most significant period of myelination occurs from mid-gestation to age 2 years.4 Before birth, myelination occurs in brain areas involved in orientation and balance. After birth, the rate of myelination of areas involved in vision and hearing reaches a peak before myelination of areas underlying language, coinciding with the emergence of these abilities.2 | Apoptosis is programmed cell death. Of all the cells that are produced in the brain, about half die through a variety of mechanisms. One of these mechanisms is programmed cell death, which is regulated primarily by neurotrophic factors, such as BDNF and IGF-1. When levels of neurotrophic factors are below a certain threshold, molecules within the cell trigger degeneration. Neuron apoptosis coincides with the period of synaptogenesis, beginning during gestation and continuing through adolescence.5 Small head size or brain volume may be caused by decreased neuron proliferation or increased apoptosis. |
| Protein-energy malnutrition | Human autopsy studies and magnetic resonance imaging studies have shown that infants with IUGR had fewer brain cells and cerebral cortical grey matter volume than normal-birth-weight infants.6,7 | A human autopsy study showed that 3–4-month-old infants with moderate malnutrition [low weight for age] had decreased dendritic span and arborization [complexity of branching projections] compared to well-nourished infants.9 Rodent models have found similar effects of early postnatal undernutrition on dendrite growth.10 | Both prenatal and postnatal undernutrition in rodents results in fewer synapses as well as synaptic structural changes.a11,12 | Adults who had been exposed to famine in utero in Holland during World War II showed increased white matter hyperintensities, shown by MRI.13 Reduced myelination has been found in animal models of IUGR6 and maternal nutrient restriction without resulting in IUGR.14 IUGR decreases IGF-1 levels and IGF-1-binding protein expression, which influence myelin production.6 | BDNF and IGF-1 levels decreased and cell death increased in the offspring of baboon mothers fed a nutrient-restricted diet during pregnancy [without IUGR].14 |
| Human autopsy studies have also shown that infants with severe acute malnutrition have fewer brain cells than well-nourished infants.8 IUGR in animals results in similar effects.6 | Other animal models have also shown that IUGR decreases IGF-1 and IGF-1-binding protein expression.6 | ||||
| Fatty acids | Neurogenesis requires the synthesis of large amounts of membrane phospholipid from fatty acids. Reduced neuron proliferation has been shown in animals with gestational DHA deficiency.15 | Arachidonic acid and docosahexaenoic acid [DHA] in membranes at synaptic sites play a role in the maturation of synapses and in neurotransmission.16 | Fatty acids are structural components of myelin. Both prenatal and postnatal fatty acid deficiency in rodents reduces the amount and alters the composition of myelin.17,18 | ||
| Iron | Iron is required for the enzyme ribonucleotide reductase that regulates central nervous system cell division.6 While gestational and neonatal iron deficiency in rodents does not affect overall brain size, a decrease in the size of the hippocampus [a subcortical structure that underlies learning and memory] has been shown.19 | Gestational and neonatal iron deficiency in rodents results in truncated dendritic branching in the hippocampus, which persists into adulthood despite iron repletion.20 | Gestational and early postnatal iron deficiency in rodents results in decreased synaptic maturity and efficacy in the hippocampus, which persists despite iron repletion.21 | Iron plays a role in myelin synthesis. In animal models, even marginal iron deficiency during prenatal and early postnatal development decreases myelin synthesis and alters myelin composition, which is not corrected with iron repletion.24 | |
| In adult rodents, iron deficiency decreases the number of dopamine D2 receptors and the density of dopamine transporter in the striatum and nucleus accumbens.22 | |||||
| In both animal models and cell culture experiments, dopamine and norepinephrine metabolism are altered by iron deficiency.23 | |||||
| Iodine and thyroid hormones | Some fetuses aborted in months 6 and 8 of gestation in an iodine-deficient area of China had lower brain weight than fetuses in an iodine-sufficient area, while some showed increased cell density. Gestational iodine deficiency in sheep and marmosets resulted in reduced brain weight and cell number, which was not corrected with iodine repletion. No effect on brain weight or cell number was found in rodents with gestational iodine deficiency, but cell migration was impaired.25 | Gestational iodine deficiency results in reduced dendritic branching in the cerebral cortex in rodents26 and in the cerebellum in sheep and marmosets.25 | Gestational iodine deficiency in sheep resulted in decreased synaptic density, which was not corrected with iodine repletion.25 | Gestational and early postnatal hypothyroidism in rodents decreases the number and density of synapses in the cerebellum, and alters neurotransmitter levels.27 | No myelination was detected in the cerebral cortex of fetuses aborted at month 8 of gestation in an iodine-deficient area of China.28 Gestational iodine deficiency in sheep and rodents reduces myelination.25 |
| Early postnatal hypothyroidism in rodents results in decreased dendritic branching in the visual and auditory cortex and cerebellum.27 | Gestational and early postnatal hypothyroidism in rodents leads to reduced myelination.27 | ||||
| Zinc | Zinc is necessary for cell division due to its role in DNA synthesis. Gestational zinc deficiency in rodents results in decreased number of cells, as reflected by total brain DNA29 and reduced regional brain mass in the cerebellum, limbic system, and cerebral cortex.6 | Gestational zinc deficiency in rodents results in reduced dendritic arborization.6 | Zinc released into synapses in the hippocampus and cerebral cortex modulates synapse function. Specifically, zinc modulates postsynaptic NMDA receptors for glutamate and inhibits GABAB receptor activation.30 | In rodent pups, zinc deficiency decreased expression of IGF-1 and growth hormone receptor genes.31 | |
| Choline | Choline is essential for stem cell proliferation and is involved in transmembrane signaling during neurogenesis.6 In rodents, gestational choline supplementation stimulates cell division.32 | The neurotransmitter acetylcholine is synthesized from choline. Gestational choline deficiency in rodents has long-term effects on cholinergic neurotransmission despite repletion.32 | Gestational choline deficiency increases the rate of apoptosis in the hippocampus in rodents.32 | ||
| B-vitamins | Before neuron proliferation begins, during weeks 2−4 of gestation, the neural tube forms, which is comprised of progenitor [stem] cells that give rise to neurons and glial cells [cells that support neurons].2 Maternal deficiency in folic acid and vitamin B12 is associated with neural tube defects, such as anencephaly and spina bifida.33 | Gestational and early postnatal vitamin B6 deficiency in rodents results in reduced dendritic branching in the neocortex and cerebellum.34,35 | Gestational and early postnatal vitamin B6 deficiency in rodents results in decreased synaptic density in the neocortex,35 reduced synaptic efficiency, particularly in NMDA receptors,36 and lowered dopamine levels and dopamine D2 receptor binding in the striatum.37 | Gestational and early postnatal vitamin B6 deficiency in rodents results in reduced myelination.38 | |
| Experience | Rodents raised in enriched environments [large enclosures with objects that allow visual and tactile stimulation] show greater brain weight and cortical thickness than rodents raised in impoverished environments [standard lab cages].39 | A human autopsy study showed that individuals with higher levels of education had more dendritic branching than those with lower education in Wernicke's area, a brain area underlying language processing.40 Rodents raised in enriched environments [filled with toys and other rodents] have more dendritic spines than those raised in less complex environments.41 | Rodents raised in enriched environments [large enclosures with objects that allow visual and tactile stimulation] show more synapses per neuron in visual and motor cortices than rodents raised in impoverished environments [standard lab cages].39 | Children raised in Romanian orphanages and then adopted into US families, thus having experienced a degree of early socioemotional deprivation, showed structural changes in white matter tracts compared to control children who had not spent any time in an orphanage.42 | |
| Practicing the piano in childhood correlated with myelination in areas underlying finger movements, as measured by fractional anistropy.43 | |||||
| An enriched rearing environment affects myelination of the corpus callosum in rodents and monkeys.44,45 |
a
The gestational period in rodents corresponds to the first half of pregnancy in humans, while the first 3 weeks after birth in rodents corresponds to the second half of pregnancy in humans.46
Abbreviations: BDNF, brain-derived neurotrophic factor; GABAB, gamma-aminobutyric acid B; IGF-1, insulin-like growth factor-1; IUGR, intrauterine growth restriction; NMDA, N-methyl-D-aspartate.
Table 1
Evidence for the role of selected nutrients and experience in five key neurodevelopmental processes
| Definition and timing | Neuron proliferation is the creation of new cells through cell division. This begins in week 7 of gestation and continues to at least 4.5 months postpartum.2 Neuron proliferation is mostly completed at birth, but neurons can be created in adulthood.3 | Axons and dendrites are branching projections that grow out from cell bodies to make connections with other cells. This process begins during gestation and continues through at least 2 years after birth. In some brain areas, axons reach their final destinations at 15 weeks gestation, in others at 32 weeks gestation. Dendrite growth begins at 15 weeks gestation and continues through the second year after birth in some brain areas.2 | Synapses are connections between axons, dendrites, and cell bodies. Synapse formation begins during gestation [around week 23] and continues throughout the lifespan. Synaptic density reaches a peak at different times in different brain areas [for example, in the visual cortex between 4 and 12 months postpartum, and in the prefrontal cortex after 15 months postpartum].3 The decrease in synaptic density that follows this peak in each area reflects synaptic pruning. Synapse overproduction is completed in the second year after birth, while synaptic pruning begins in the first year after birth and continues through adolescence.2 | Myelin is white, fatty matter that covers axons and accelerates the speed of nerve impulses traveling from one cell to another. Myelination begins as early as 12–14 weeks of gestation in the spinal cord and continues until adulthood. The most significant period of myelination occurs from mid-gestation to age 2 years.4 Before birth, myelination occurs in brain areas involved in orientation and balance. After birth, the rate of myelination of areas involved in vision and hearing reaches a peak before myelination of areas underlying language, coinciding with the emergence of these abilities.2 | Apoptosis is programmed cell death. Of all the cells that are produced in the brain, about half die through a variety of mechanisms. One of these mechanisms is programmed cell death, which is regulated primarily by neurotrophic factors, such as BDNF and IGF-1. When levels of neurotrophic factors are below a certain threshold, molecules within the cell trigger degeneration. Neuron apoptosis coincides with the period of synaptogenesis, beginning during gestation and continuing through adolescence.5 Small head size or brain volume may be caused by decreased neuron proliferation or increased apoptosis. |
| Protein-energy malnutrition | Human autopsy studies and magnetic resonance imaging studies have shown that infants with IUGR had fewer brain cells and cerebral cortical grey matter volume than normal-birth-weight infants.6,7 | A human autopsy study showed that 3–4-month-old infants with moderate malnutrition [low weight for age] had decreased dendritic span and arborization [complexity of branching projections] compared to well-nourished infants.9 Rodent models have found similar effects of early postnatal undernutrition on dendrite growth.10 | Both prenatal and postnatal undernutrition in rodents results in fewer synapses as well as synaptic structural changes.a11,12 | Adults who had been exposed to famine in utero in Holland during World War II showed increased white matter hyperintensities, shown by MRI.13 Reduced myelination has been found in animal models of IUGR6 and maternal nutrient restriction without resulting in IUGR.14 IUGR decreases IGF-1 levels and IGF-1-binding protein expression, which influence myelin production.6 | BDNF and IGF-1 levels decreased and cell death increased in the offspring of baboon mothers fed a nutrient-restricted diet during pregnancy [without IUGR].14 |
| Human autopsy studies have also shown that infants with severe acute malnutrition have fewer brain cells than well-nourished infants.8 IUGR in animals results in similar effects.6 | Other animal models have also shown that IUGR decreases IGF-1 and IGF-1-binding protein expression.6 | ||||
| Fatty acids | Neurogenesis requires the synthesis of large amounts of membrane phospholipid from fatty acids. Reduced neuron proliferation has been shown in animals with gestational DHA deficiency.15 | Arachidonic acid and docosahexaenoic acid [DHA] in membranes at synaptic sites play a role in the maturation of synapses and in neurotransmission.16 | Fatty acids are structural components of myelin. Both prenatal and postnatal fatty acid deficiency in rodents reduces the amount and alters the composition of myelin.17,18 | ||
| Iron | Iron is required for the enzyme ribonucleotide reductase that regulates central nervous system cell division.6 While gestational and neonatal iron deficiency in rodents does not affect overall brain size, a decrease in the size of the hippocampus [a subcortical structure that underlies learning and memory] has been shown.19 | Gestational and neonatal iron deficiency in rodents results in truncated dendritic branching in the hippocampus, which persists into adulthood despite iron repletion.20 | Gestational and early postnatal iron deficiency in rodents results in decreased synaptic maturity and efficacy in the hippocampus, which persists despite iron repletion.21 | Iron plays a role in myelin synthesis. In animal models, even marginal iron deficiency during prenatal and early postnatal development decreases myelin synthesis and alters myelin composition, which is not corrected with iron repletion.24 | |
| In adult rodents, iron deficiency decreases the number of dopamine D2 receptors and the density of dopamine transporter in the striatum and nucleus accumbens.22 | |||||
| In both animal models and cell culture experiments, dopamine and norepinephrine metabolism are altered by iron deficiency.23 | |||||
| Iodine and thyroid hormones | Some fetuses aborted in months 6 and 8 of gestation in an iodine-deficient area of China had lower brain weight than fetuses in an iodine-sufficient area, while some showed increased cell density. Gestational iodine deficiency in sheep and marmosets resulted in reduced brain weight and cell number, which was not corrected with iodine repletion. No effect on brain weight or cell number was found in rodents with gestational iodine deficiency, but cell migration was impaired.25 | Gestational iodine deficiency results in reduced dendritic branching in the cerebral cortex in rodents26 and in the cerebellum in sheep and marmosets.25 | Gestational iodine deficiency in sheep resulted in decreased synaptic density, which was not corrected with iodine repletion.25 | Gestational and early postnatal hypothyroidism in rodents decreases the number and density of synapses in the cerebellum, and alters neurotransmitter levels.27 | No myelination was detected in the cerebral cortex of fetuses aborted at month 8 of gestation in an iodine-deficient area of China.28 Gestational iodine deficiency in sheep and rodents reduces myelination.25 |
| Early postnatal hypothyroidism in rodents results in decreased dendritic branching in the visual and auditory cortex and cerebellum.27 | Gestational and early postnatal hypothyroidism in rodents leads to reduced myelination.27 | ||||
| Zinc | Zinc is necessary for cell division due to its role in DNA synthesis. Gestational zinc deficiency in rodents results in decreased number of cells, as reflected by total brain DNA29 and reduced regional brain mass in the cerebellum, limbic system, and cerebral cortex.6 | Gestational zinc deficiency in rodents results in reduced dendritic arborization.6 | Zinc released into synapses in the hippocampus and cerebral cortex modulates synapse function. Specifically, zinc modulates postsynaptic NMDA receptors for glutamate and inhibits GABAB receptor activation.30 | In rodent pups, zinc deficiency decreased expression of IGF-1 and growth hormone receptor genes.31 | |
| Choline | Choline is essential for stem cell proliferation and is involved in transmembrane signaling during neurogenesis.6 In rodents, gestational choline supplementation stimulates cell division.32 | The neurotransmitter acetylcholine is synthesized from choline. Gestational choline deficiency in rodents has long-term effects on cholinergic neurotransmission despite repletion.32 | Gestational choline deficiency increases the rate of apoptosis in the hippocampus in rodents.32 | ||
| B-vitamins | Before neuron proliferation begins, during weeks 2−4 of gestation, the neural tube forms, which is comprised of progenitor [stem] cells that give rise to neurons and glial cells [cells that support neurons].2 Maternal deficiency in folic acid and vitamin B12 is associated with neural tube defects, such as anencephaly and spina bifida.33 | Gestational and early postnatal vitamin B6 deficiency in rodents results in reduced dendritic branching in the neocortex and cerebellum.34,35 | Gestational and early postnatal vitamin B6 deficiency in rodents results in decreased synaptic density in the neocortex,35 reduced synaptic efficiency, particularly in NMDA receptors,36 and lowered dopamine levels and dopamine D2 receptor binding in the striatum.37 | Gestational and early postnatal vitamin B6 deficiency in rodents results in reduced myelination.38 | |
| Experience | Rodents raised in enriched environments [large enclosures with objects that allow visual and tactile stimulation] show greater brain weight and cortical thickness than rodents raised in impoverished environments [standard lab cages].39 | A human autopsy study showed that individuals with higher levels of education had more dendritic branching than those with lower education in Wernicke's area, a brain area underlying language processing.40 Rodents raised in enriched environments [filled with toys and other rodents] have more dendritic spines than those raised in less complex environments.41 | Rodents raised in enriched environments [large enclosures with objects that allow visual and tactile stimulation] show more synapses per neuron in visual and motor cortices than rodents raised in impoverished environments [standard lab cages].39 | Children raised in Romanian orphanages and then adopted into US families, thus having experienced a degree of early socioemotional deprivation, showed structural changes in white matter tracts compared to control children who had not spent any time in an orphanage.42 | |
| Practicing the piano in childhood correlated with myelination in areas underlying finger movements, as measured by fractional anistropy.43 | |||||
| An enriched rearing environment affects myelination of the corpus callosum in rodents and monkeys.44,45 |
| Definition and timing | Neuron proliferation is the creation of new cells through cell division. This begins in week 7 of gestation and continues to at least 4.5 months postpartum.2 Neuron proliferation is mostly completed at birth, but neurons can be created in adulthood.3 | Axons and dendrites are branching projections that grow out from cell bodies to make connections with other cells. This process begins during gestation and continues through at least 2 years after birth. In some brain areas, axons reach their final destinations at 15 weeks gestation, in others at 32 weeks gestation. Dendrite growth begins at 15 weeks gestation and continues through the second year after birth in some brain areas.2 | Synapses are connections between axons, dendrites, and cell bodies. Synapse formation begins during gestation [around week 23] and continues throughout the lifespan. Synaptic density reaches a peak at different times in different brain areas [for example, in the visual cortex between 4 and 12 months postpartum, and in the prefrontal cortex after 15 months postpartum].3 The decrease in synaptic density that follows this peak in each area reflects synaptic pruning. Synapse overproduction is completed in the second year after birth, while synaptic pruning begins in the first year after birth and continues through adolescence.2 | Myelin is white, fatty matter that covers axons and accelerates the speed of nerve impulses traveling from one cell to another. Myelination begins as early as 12–14 weeks of gestation in the spinal cord and continues until adulthood. The most significant period of myelination occurs from mid-gestation to age 2 years.4 Before birth, myelination occurs in brain areas involved in orientation and balance. After birth, the rate of myelination of areas involved in vision and hearing reaches a peak before myelination of areas underlying language, coinciding with the emergence of these abilities.2 | Apoptosis is programmed cell death. Of all the cells that are produced in the brain, about half die through a variety of mechanisms. One of these mechanisms is programmed cell death, which is regulated primarily by neurotrophic factors, such as BDNF and IGF-1. When levels of neurotrophic factors are below a certain threshold, molecules within the cell trigger degeneration. Neuron apoptosis coincides with the period of synaptogenesis, beginning during gestation and continuing through adolescence.5 Small head size or brain volume may be caused by decreased neuron proliferation or increased apoptosis. |
| Protein-energy malnutrition | Human autopsy studies and magnetic resonance imaging studies have shown that infants with IUGR had fewer brain cells and cerebral cortical grey matter volume than normal-birth-weight infants.6,7 | A human autopsy study showed that 3–4-month-old infants with moderate malnutrition [low weight for age] had decreased dendritic span and arborization [complexity of branching projections] compared to well-nourished infants.9 Rodent models have found similar effects of early postnatal undernutrition on dendrite growth.10 | Both prenatal and postnatal undernutrition in rodents results in fewer synapses as well as synaptic structural changes.a11,12 | Adults who had been exposed to famine in utero in Holland during World War II showed increased white matter hyperintensities, shown by MRI.13 Reduced myelination has been found in animal models of IUGR6 and maternal nutrient restriction without resulting in IUGR.14 IUGR decreases IGF-1 levels and IGF-1-binding protein expression, which influence myelin production.6 | BDNF and IGF-1 levels decreased and cell death increased in the offspring of baboon mothers fed a nutrient-restricted diet during pregnancy [without IUGR].14 |
| Human autopsy studies have also shown that infants with severe acute malnutrition have fewer brain cells than well-nourished infants.8 IUGR in animals results in similar effects.6 | Other animal models have also shown that IUGR decreases IGF-1 and IGF-1-binding protein expression.6 | ||||
| Fatty acids | Neurogenesis requires the synthesis of large amounts of membrane phospholipid from fatty acids. Reduced neuron proliferation has been shown in animals with gestational DHA deficiency.15 | Arachidonic acid and docosahexaenoic acid [DHA] in membranes at synaptic sites play a role in the maturation of synapses and in neurotransmission.16 | Fatty acids are structural components of myelin. Both prenatal and postnatal fatty acid deficiency in rodents reduces the amount and alters the composition of myelin.17,18 | ||
| Iron | Iron is required for the enzyme ribonucleotide reductase that regulates central nervous system cell division.6 While gestational and neonatal iron deficiency in rodents does not affect overall brain size, a decrease in the size of the hippocampus [a subcortical structure that underlies learning and memory] has been shown.19 | Gestational and neonatal iron deficiency in rodents results in truncated dendritic branching in the hippocampus, which persists into adulthood despite iron repletion.20 | Gestational and early postnatal iron deficiency in rodents results in decreased synaptic maturity and efficacy in the hippocampus, which persists despite iron repletion.21 | Iron plays a role in myelin synthesis. In animal models, even marginal iron deficiency during prenatal and early postnatal development decreases myelin synthesis and alters myelin composition, which is not corrected with iron repletion.24 | |
| In adult rodents, iron deficiency decreases the number of dopamine D2 receptors and the density of dopamine transporter in the striatum and nucleus accumbens.22 | |||||
| In both animal models and cell culture experiments, dopamine and norepinephrine metabolism are altered by iron deficiency.23 | |||||
| Iodine and thyroid hormones | Some fetuses aborted in months 6 and 8 of gestation in an iodine-deficient area of China had lower brain weight than fetuses in an iodine-sufficient area, while some showed increased cell density. Gestational iodine deficiency in sheep and marmosets resulted in reduced brain weight and cell number, which was not corrected with iodine repletion. No effect on brain weight or cell number was found in rodents with gestational iodine deficiency, but cell migration was impaired.25 | Gestational iodine deficiency results in reduced dendritic branching in the cerebral cortex in rodents26 and in the cerebellum in sheep and marmosets.25 | Gestational iodine deficiency in sheep resulted in decreased synaptic density, which was not corrected with iodine repletion.25 | Gestational and early postnatal hypothyroidism in rodents decreases the number and density of synapses in the cerebellum, and alters neurotransmitter levels.27 | No myelination was detected in the cerebral cortex of fetuses aborted at month 8 of gestation in an iodine-deficient area of China.28 Gestational iodine deficiency in sheep and rodents reduces myelination.25 |
| Early postnatal hypothyroidism in rodents results in decreased dendritic branching in the visual and auditory cortex and cerebellum.27 | Gestational and early postnatal hypothyroidism in rodents leads to reduced myelination.27 | ||||
| Zinc | Zinc is necessary for cell division due to its role in DNA synthesis. Gestational zinc deficiency in rodents results in decreased number of cells, as reflected by total brain DNA29 and reduced regional brain mass in the cerebellum, limbic system, and cerebral cortex.6 | Gestational zinc deficiency in rodents results in reduced dendritic arborization.6 | Zinc released into synapses in the hippocampus and cerebral cortex modulates synapse function. Specifically, zinc modulates postsynaptic NMDA receptors for glutamate and inhibits GABAB receptor activation.30 | In rodent pups, zinc deficiency decreased expression of IGF-1 and growth hormone receptor genes.31 | |
| Choline | Choline is essential for stem cell proliferation and is involved in transmembrane signaling during neurogenesis.6 In rodents, gestational choline supplementation stimulates cell division.32 | The neurotransmitter acetylcholine is synthesized from choline. Gestational choline deficiency in rodents has long-term effects on cholinergic neurotransmission despite repletion.32 | Gestational choline deficiency increases the rate of apoptosis in the hippocampus in rodents.32 | ||
| B-vitamins | Before neuron proliferation begins, during weeks 2−4 of gestation, the neural tube forms, which is comprised of progenitor [stem] cells that give rise to neurons and glial cells [cells that support neurons].2 Maternal deficiency in folic acid and vitamin B12 is associated with neural tube defects, such as anencephaly and spina bifida.33 | Gestational and early postnatal vitamin B6 deficiency in rodents results in reduced dendritic branching in the neocortex and cerebellum.34,35 | Gestational and early postnatal vitamin B6 deficiency in rodents results in decreased synaptic density in the neocortex,35 reduced synaptic efficiency, particularly in NMDA receptors,36 and lowered dopamine levels and dopamine D2 receptor binding in the striatum.37 | Gestational and early postnatal vitamin B6 deficiency in rodents results in reduced myelination.38 | |
| Experience | Rodents raised in enriched environments [large enclosures with objects that allow visual and tactile stimulation] show greater brain weight and cortical thickness than rodents raised in impoverished environments [standard lab cages].39 | A human autopsy study showed that individuals with higher levels of education had more dendritic branching than those with lower education in Wernicke's area, a brain area underlying language processing.40 Rodents raised in enriched environments [filled with toys and other rodents] have more dendritic spines than those raised in less complex environments.41 | Rodents raised in enriched environments [large enclosures with objects that allow visual and tactile stimulation] show more synapses per neuron in visual and motor cortices than rodents raised in impoverished environments [standard lab cages].39 | Children raised in Romanian orphanages and then adopted into US families, thus having experienced a degree of early socioemotional deprivation, showed structural changes in white matter tracts compared to control children who had not spent any time in an orphanage.42 | |
| Practicing the piano in childhood correlated with myelination in areas underlying finger movements, as measured by fractional anistropy.43 | |||||
| An enriched rearing environment affects myelination of the corpus callosum in rodents and monkeys.44,45 |
a
The gestational period in rodents corresponds to the first half of pregnancy in humans, while the first 3 weeks after birth in rodents corresponds to the second half of pregnancy in humans.46
Abbreviations: BDNF, brain-derived neurotrophic factor; GABAB, gamma-aminobutyric acid B; IGF-1, insulin-like growth factor-1; IUGR, intrauterine growth restriction; NMDA, N-methyl-D-aspartate.
Evidence from animal models of nutrient deficiency, and some evidence from human studies, clearly shows that many nutrients are necessary for brain development. Table 1 presents evidence for the effect of specific nutrient deficiencies during early development on five key neurodevelopmental processes: 1] neuron proliferation, 2] axon and dendrite growth, 3] synapse formation, pruning, and function, 4] myelination, and 5] neuron apoptosis [programmed cell death]. Table 1 focuses on nutrients that have been studied in human as well as animal studies. Other nutrients, such as copper, which is also important for some of these neurodevelopmental processes, are not included since rigorous studies in human populations and intervention studies have not yet been conducted.
Although the necessity of nutrients for brain development is evident, the extent to which nutrient deprivation during gestation and infancy results in long-term effects on brain function in free-living human populations is not yet clear. The actual impact depends on several factors, including 1] the child's experience and input from the environment, 2] the timing of nutrient deprivation, 3] the degree of nutrient deficiency, and 4] the possibility of recovery. Each of these factors is discussed in the following sections, followed by a brief discussion of methodological factors that can also influence the results of nutrition studies.
Factors Influencing the Impact of Undernutrition
Experience and input from the environment
Brain development is affected by experience. Two types of processes are described as “experience-expectant” and “experience-dependent.”41 In experience-expectant processes, the brain relies on specific input for normal development. For example, the brain expects visual input through the optic nerve for normal development of the visual cortex.41 The absence of these expected experiences impairs the neurodevelopmental processes that depend on them. These experience-expectant processes also depend on other types of sensory stimulation [e.g., auditory and tactile] and occur early in life. In contrast, “experience-dependent” processes refer to the way the brain organizes itself in response to an individual's experiences and acquired skills, which is a process that continues throughout the lifespan. For example, a neuroimaging study demonstrated that the rear hippocampus, a part of the brain that underlies spatial memory, increased in volume as London taxi-driver trainees learned the layout of the city streets.47 While experience-expectant mechanisms refer to features of the environment that are [or should be] universal, experience-dependent mechanisms refer to aspects of the environment that are unique to the individual. These latter processes enable individuals to adapt to and thrive in their specific culture and environment.
Adequate nutrition can be considered an aspect of the environment that is expected by the brain for normal development.48 An environment with poor quality and variety of sensory and social input impairs some of the same neurodevelopmental processes as nutrient deprivation during early development, including the complexity of dendritic branching and synaptic density [Table 1]. The parallel influences of nutrient deficiency and stimulation from the environment on brain development may operate in several ways: additive effects, interacting effects, and mediating effects, all of which have been demonstrated in empirical studies. These are depicted in Figure 1 and discussed in greater detail below.
Figure 1
Three hypothetical scenarios in which the effects of undernutrition and a poor-quality environment may show additive or interacting effects on children's motor, cognitive, and socioemotional development. A Additive effects of undernutrition and poor-quality environment. B An enriched environment protects children from negative effects of undernutrition. C Nutrition intervention only affects children who have adequate stimulation, or stimulation intervention only affects children who have adequate nutrition.
Additive effects
Nutrient deficiency and experiential input from the environment may have independent additive effects on brain development. In this case, in an at-risk population, one would expect children with both risk factors [nutrient deficiency and low stimulation] to perform at low levels, children with one risk factor [nutrient deficiency or low stimulation] to perform at average levels, and children with neither risk factor [sufficient nutrition and high stimulation] to perform at high levels in cognitive, motor, and socioemotional development. This pattern is shown in Figure 1a. In support of this hypothesis, several studies have shown that nutritional supplementation and psychosocial stimulation together result in greater improvements in child development than either intervention alone.49,50 In these studies, psychosocial stimulation consisted of periodic home visits during which community workers facilitated play sessions with mothers and children. The community workers conducted activities such as demonstrating play with homemade toys, emphasizing the quality of the verbal interactions between mothers and children, and teaching concepts such as color, shape, size, and number. Children in Costa Rica showed similar additive effects of iron-deficiency anemia in infancy and low socioeconomic status on cognitive scores at school age.51
Interacting effects
Alternatively, nutrient deficiency or intervention may affect some children but not others, depending on the amount and quality of stimulation they receive. For example, in Chile, low-birth-weight infants born into families with high socioeconomic status were at lower risk for poor developmental outcomes than those born into disadvantaged environments.52 Similarly, in 6−8-year-old children in Vietnam, nutritional status was related to cognitive scores among children who did not participate in a preschool program at age 3−4 years, but not among those who did.53 Thus, in some cases, stimulation from the environment can protect children from negative effects of undernutrition on development. This possibility is shown in Figure 1b. Conversely, undernourished children from disadvantaged homes where protective factors are lacking may show more of a developmental response to nutrition and other forms of interventions. For example, in Guatemala, the effect of a supplementary protein/energy drink on infant and preschool development was greatest among families of low socioeconomic status.54 In Chile, 1 year of weekly home visits providing psychosocial stimulation increased cognitive and socioemotional scores in infants with iron-deficiency anemia [IDA], but not in infants without IDA.55
Another way that nutrition and stimulation may interact is that nutritional supplementation may only positively affect development among children who receive a certain amount of stimulation from the environment. If children do not receive any stimulation, improving nutrition alone may be insufficient to improve brain development. For example, in Jamaica, infants between the ages of 9 and 30 months who participated in a psychosocial stimulation intervention benefited from zinc supplementation, while those who did not receive psychosocial stimulation did not show any developmental benefit from supplementary zinc.56 It is also possible that an intervention providing psychosocial stimulation may only benefit children who are adequately nourished. In an animal model of maternal choline deficiency, 7-month-old rodents were exposed to an environmental enrichment experience by being allowed to explore a maze once daily for 12 days. Rodents whose mothers had been given choline during gestation showed increased neurogenesis in the hippocampus through the enriching experience, while rodents whose mothers had been deprived of choline during gestation did not show altered neurogenesis.57 This type of pattern is illustrated in Figure 1c.
Mediating effects
Finally, improving nutritional status may actually improve children's experiences and the stimulation they receive from the environment. Undernutrition affects physical growth, physical activity, and motor development, which may, in turn, influence brain development through two pathways. The first pathway is through caregiver behavior and the second is through child exploration of the environment54,58 [see Figure 2]. First, caregivers may treat children who are small for their age as younger than they actually are, and thus not provide age-appropriate stimulation, which could result in altered brain development. Also, undernourished children may be frequently ill and therefore fussy, irritable, and withdrawn, leading caregivers to treat them more negatively than they would treat a happy, healthy child. Reduced activity due to undernutrition may limit the child's exploration of the environment and initiation of caregiver interactions, which could also lead to poor brain development.59 Some evidence suggests that these mechanisms contribute to delayed motor and cognitive development in infants and children with IDA.60,61 However, in stunted Jamaican infants, nutritional supplementation affected cognitive development but not activity levels, and activity and development were not related to each other, suggesting that this mechanism did not mediate the effect of nutrition on cognitive development in this cohort.62
Figure 2
Hypothetical scenario in which the child's experience acts as a mediator between nutritional status and motor, cognitive, and socioemotional development. Adapted from Levitsky and Barnes58 and Pollitt.54
Few studies have examined the potential additive, interacting, and mediating effects of nutrition and experiential input from the environment on child motor, cognitive, and socioemotional development. Studies that have tested all of these in a systematic way could not be located in the existing literature. In future research, datasets that allow the testing of each of these hypotheses are needed.
Timing of nutrient deprivation or supplementation
Nutrient deficiency is more likely to impair brain development if the deficiency occurs during a time period when the need for that nutrient for neurodevelopment is high. Various nutrients are necessary for specific neurodevelopmental processes. Each process occurs in different, overlapping time periods in different brain areas. The timing of five key neurodevelopmental processes is presented in the first row of Table 1. Drawing links between specific nutrients, specific neurodevelopmental processes, and the time period of deprivation or supplementation allows specific hypotheses to be made concerning the effect of nutrient deprivation or supplementation on brain development.
For example, myelination of the brainstem auditory pathway occurs from week 26 of gestation until at least 1 year after birth.63 Fatty acids such as docosahexaeonic acid [DHA] are necessary for myelination. This leads to the hypothesis that supplementation with DHA in the third trimester and the first year after birth may improve myelination of this auditory pathway. The latency of auditory-evoked potentials, which measure electrical activity in response to an auditory stimulus through electrodes placed on the scalp, is thought to reflect myelination, among other physiological aspects of brain function.64 In support of the effect of DHA on myelination of the brainstem auditory pathway during the first few months after birth, a study in Turkey demonstrated that infants fed a formula containing DHA showed more rapid brainstem auditory-evoked potentials at age 16 weeks than infants fed a formula without DHA.65 Future studies that examine precise hypotheses related to specific nutrients, neurodevelopmental processes, timing, and brain areas are needed to clarify the relationship between nutrition and brain development and its mechanisms. For a more complete discussion of the timing of neurodevelopmental processes and implications for measurement see Georgieff66 and Wachs et al.67
Degree of nutrient deficiency
Much evidence shows that brain development may be compromised when nutrient deficiency is severe to moderate but spared when deficiency is mild to moderate. A number of homeostatic mechanisms protect the developing fetus and the developing brain from nutrient deficiency to a certain degree. For example, in the case of placental insufficiency, when insufficient nutrients and oxygen are available, fetal cardiac output is redistributed such that blood flow to the peripheral tissues decreases and blood flow to the brain, adrenal glands, and heart increases. This leads to brain sparing, or the sparing of brain growth even when overall fetal growth is reduced.68 Another mechanism that protects the fetus from iron deficiency to a certain degree is the increased transfer of iron across the placenta as maternal levels decrease.69 For each nutrient, there is likely to be a threshold at which deficiency results in impairment for the child. Exactly where this line is drawn is an important question which must be answered for each nutrient individually.
Several studies have shown that the effect of nutritional supplementation on brain development depends on initial nutritional status. For example, in Bangladesh and Indonesia, a positive effect of maternal multiple micronutrient supplementation during pregnancy and postpartum on child motor and cognitive development was found only in children of undernourished mothers.70,71 Similarly, in Chile, infants with low hemoglobin concentration at age 6 months showed improved cognition at age 10 years if they had been fed iron-fortified formula [compared to low-iron formula] during infancy, whereas children with high hemoglobin concentration at age 6 months performed better in cognitive tasks at age 10 years if they had received low-iron formula.72 In summary, greater severity of nutritional deficiency increases both the likelihood of negative effects on brain development and the likelihood of positively responding to nutritional supplementation.
Possibility of recovery
Even if the timing and the degree of nutrient deficiency are sufficient to alter brain development, one important question is whether these changes can be subsequently corrected. If not, children undernourished in early life would show permanent developmental deficits. On the other hand, if some or all of these structural alterations can be corrected, children could partly or fully recover cognitive ability.
The brain's potential for recovery from early damage has been widely studied in the context of neurological injury during development. When a certain part of the brain is damaged during early life, recovery happens in three ways, depending on the timing of the injury and subsequent experience. First, there are changes in the organization of the remaining intact circuits in the brain that were left uninjured, involving the generation of new synapses in existing pathways. Second, new circuitry that did not exist before the injury develops. Third, neurons and glia are generated to replace the injured neurons and glia.73 In the case of brain alterations caused by nutrient deficiency, recovery is plausible if nutrients become available during the time that the affected growth process is still occurring. In addition to nutrient repletion, enhanced sensory, linguistic, and social interactions may also facilitate recovery.
Data from a group of Korean orphans adopted by middle-class American families provided an opportunity to investigate the possibility of recovery. Children who were undernourished at the time of adoption [before age 2 years] did not score below the normal range on IQ tests at school age, but their scores were lower than those of Korean adoptees who had not been undernourished in infancy.74 In addition, children adopted after age 2 years had lower IQ scores than those adopted before age 2 years, suggesting that improved conditions earlier rather than later in childhood provide a greater benefit.75
Other investigators have studied adults who were born during a period of famine in Holland during World War II when strict food rations were imposed on the entire Dutch population, including pregnant women. Children born during this period experienced nutrient deprivation in utero but adequate nutrition and health care thereafter. At age 19 years, their average IQ did not differ from that of a group whose mothers did not experience famine during pregnancy.76 However, adults exposed to this famine in utero had increased risk of diagnosis of schizophrenia77 and antisocial personality disorder,78 as well as admittance to an addiction treatment program.79 Together, this evidence suggests that some, but not all, of the negative effects of early undernutrition on brain development can be reversed through subsequent improvement in nutrition, health care, and enriched environments.
In these studies, the role of improved nutrition and the role of stimulation from the environment in recovery cannot be distinguished. Other evidence suggests that both of these can contribute to cognitive recovery after early undernutrition. In a large cohort of Peruvian children [n = 1,674], children who had been stunted before age 18 months but who were not stunted at age 4−6 years performed as well as children who had never been stunted in vocabulary and quantitative tests, while children who did not experience catch-up growth scored significantly lower.80 In other studies, providing cognitive stimulation to children who suffered from an episode of severe acute malnutrition or IDA in early life improved mental and motor development.55,81 This type of evidence has led the World Health Organization to recommend structured activities to promote cognitive development as a component of the treatment of early childhood malnutrition, in addition to nutrition and healthcare.82
Methodological factors
The selection of assessment tools and the age of assessment can also influence whether effects are found in nutrition studies. Global measures, such as the Bayley Scales of Infant Development [BSID] or IQ tests, are widely used but may be less sensitive to nutritional deficiency than tests of specific cognitive abilities.83,–85 In addition, using a test created in a high-income country in a low-income country without adaptation can lead to systematic bias.86,87 For a more complete discussion of assessing cognitive abilities in nutrition studies, see Isaacs and Oates.84
Detecting the effects of early nutrient deficiency can also depend on the age of cognitive assessment. For example, a group of children who experienced thiamine deficiency in infancy did not show neurological symptoms at the time of deficiency, but showed language impairment at age 5–7 years.88 Similarly, in a randomized controlled trial, infants who received formula containing certain fatty acids [docosahexaenoic acid and arachidonic acid] showed higher vocabulary and IQ scores at age 5–6 years compared to infants who received formula without these fatty acids, even though they did not differ in vocabulary or BSID scores at age 18 months.85 These examples show that long-term effects may be found even when immediate effects of early nutritional deficiency are not apparent.
In summary, the long-term effect of nutritional deficiency on brain development depends on the timing and degree of deficiency, as well as the quality of the child's environment. Recovery is possible with nutrient repletion during a time period when the affected neurodevelopmental process is ongoing and with enhanced interaction with caregivers and other aspects of the environment.
Brief Review of Human Studies
As shown in Table 1, research in animals has demonstrated the effects of many specific nutrient deficiencies on the development of brain structure and function. However, studies examining the effect of mild to moderate undernutrition on brain development in free-living mothers and children have largely shown mixed or inconclusive results. The factors discussed such as the timing and degree of deficiency and interactions with the amount of stimulation children receive may account for some of these mixed results. In addition, in many studies, undernutrition is confounded by other factors such as poverty, unstimulating environments, little maternal education, poor healthcare, and preterm birth, which make it difficult to isolate the effects of nutrition. To do this, randomized controlled trials are needed, but few of these specifically examining neurobehavioral outcomes have been conducted. The following sections briefly review studies of the long-term consequences of undernutrition in early life, food and protein/energy supplementation, breastfeeding practices, essential fatty acids, and certain specific micronutrients, with a focus on studies from low- and middle-income countries.
Long-term consequences of undernutrition in early life
Many studies have compared school-age children who had suffered from an episode of severe acute malnutrition in the first few years of life to matched controls or siblings who had not. These studies generally showed that those who had suffered from early malnutrition had poorer IQ levels, cognitive function, and school achievement, as well as greater behavioral problems.89 A recent study in Barbados showed that adults who had suffered from an episode of moderate to severe malnutrition in the first year of life showed more attention problems90 and lower social status and standard of living91 than matched controls, even after 37–43 years.
Chronic malnutrition, as measured by physical growth that is far below average for a child's age, is also associated with reduced cognitive and motor development. From the first year of life through school age, children who are short for their age [stunted] or underweight for their age score lower than their normal-sized peers [on average] in cognitive and motor tasks and in school achievement. Longitudinal studies that have followed children from infancy throughout childhood have also consistently shown that children who became stunted [height for age < −2 SD below norm values] before 2 years of age continued to show deficits in cognition and school achievement from the age of 5 years to adolescence.92
Growth faltering can begin before birth, and the evidence indicates that being born small for gestational age is associated with mild to moderately low performance in school during childhood and adolescence, and with lower psychological and intellectual performance in young adulthood.93 However, recent studies in low- and middle-income countries that have examined the relationship between low birth weight [