SEARCH

SEARCH BY CITATION

Philip James: This discussion will be wide-ranging and will include new input and will not follow the sequence of paper presentations.

THE RECEPTOR BIOLOGY OF THE BRAIN

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Philip James: First, let us consider receptor biology. Ole Petter emphasized that not only should we think about the marked capacity for adaptation in different regions of the brain in response to stimuli but also the question of the extraordinary receptor turnover with receptor recycling and receptor transfer to other parts of the cell and even transmission across synapses into adjacent cells with synaptic renewal. This then raises the issue of the control of receptor shuttling and the physical and physiological responses involved in learning.

Christian Drevon: The LDL [low-density lipoprotein] receptor on average has a half-life in the membrane of about 6 minutes and then moves in from the coated pit and ends up in the endosome. It is then recycled back to the membrane 12–15 times before it is degraded, so the issue is whether there are comparable data on the glutamate receptor in synapses and whether this same phenomenon is seen in the CNS [central nervous system].

Ole Petter Ottersen: I confess the field has not progressed to the same level of understanding as for LDL receptors. The very new data I presented suggest that there is an endocytotic zone that is very close to the post-synaptic density. In fact, now we know that the Homer family of molecular scaffolds provides spatial organization to regulate post-synaptic signaling cascades in a very organized manner, very close to where the receptors are inserted. These scaffolding proteins are the basic molecular players that are responsible for endocytosis, and the idea now is that this endocytotic process sets the level or the number of alpha-amino-3-hydroxy-5-methyl-4-isoxazole-proprionic acid (AMPA) receptors that are expressed in the plasma membrane. What sets the gain in the system is not known – we do not know exactly the factors controlling this endocytotic process and the receptor density in the plasma membrane. However, we do know that the number of AMPA receptors, crucial for glutamate function in the membrane, is controlled by long-term potentiation (LTP) and long-term depression (LTD). These are known physiological processes that control the level of AMPA receptors in the post-synaptic density. While cortical stimulation alone results in LTD, the combination with dopamine switches LTD to LTP, which is known as dopamine-dependent plasticity.

Ricardo Uauy: I think we have also to consider the lipid components, which are modulated by diet and modify receptor activity and turnover. The LDL receptors have been well studied at the level of both gene expression and receptor activity, as have the functional implications in terms of cholesterol synthesis and the SNARE [soluble N-ethylmaleimide-sensitive-factor attachment protein receptor] system and HMG-CoA [3-hydroxy-3methyl-glutaryl-CoA] reductase activity. In the case of the nervous system, we also know that the reuptake of neurotransmitters, which is receptor mediated with SNARE proteins, is also modulated by diet. So, the lipids are not just sitting there; there are lipid-protein interactions that define both turnover and activity and are highly relevant for normal homeostasis. They are even more important during development and aging.

Cutberto Garza: Are there any data on the age dependence of receptor turnover or recycling? Is it a function of senescence or is it the same throughout the lifespan?

Mats Rudling: I think when it comes to the LDL receptor, if you try to calculate turnover of receptors, then there are very different results depending on which tissue or cell type you are considering.

Greg Cole: The beta species influences the rate of receptor internalization for both the NMDA [N-methyl-D-aspartate] receptor and the AMPA receptor. These mobile receptors are anchored by adaptor proteins that are regulated by palmitoylation. So, nutritional interventions with things like docosahexaenoic acid (DHA) influence the palmitoylation rates of ESD 95 and the other proteins that anchor the receptor complexes.

Tom Kirkwood: Ole Petter has shown us how technological advances are beginning to open up ways of addressing questions that were not previously possible. But most of the data are from in vitro studies, and what we really want to know is what happens in vivo. When we look at the dynamics of these systems, how are they affected by nutrition? In the brain during sleep turnover may be very different from that during the waking hours. So, how can we begin to understand how these things work in real time in situ? It is not going to be easy. The second technological challenge is to understand how these things work at the single-cell level so that we can define changes with aging. We tend to treat aging as if it were some homogeneous process, but very often what happens is that most cells do not change at all but an increasing fraction of the cells are not functioning well. Therefore, in monitoring average changes over time, the average set of values does not reflect the real changes within specific cells within the aging brain.

Ole Petter Ottersen: First, one has to be very careful when extrapolating from the in vitro to the in vivo situation, but we do now have the technology that can be used to address this. We can study expressed labeled proteins, for example, AMPA receptors, in the intact animal using multiphoton imaging to see how these proteins behave, but it is technically extremely demanding. However, we have to be extremely careful when extrapolating from the number of spines on a brain cell to performance because it is not clear that the more spines you have, the better the function. In extreme conditions such as Alzheimer's disease, where you have the extreme pruning of spines, there is no doubt that to endow the cells with an increased number of synapses would help, but when considering a more or less normal functioning organism, then an increase in spine number may not necessarily improve function. There is now much discussion as to whether learning in fact will necessarily bring about an increased number of spines. It could be that it is the quality of the individual spines that really matters.

SPECIES DIFFERENCES

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Ole Petter Ottersen: To what extent can we extrapolate from mice and rats to humans? It is not only the difference in volume and surface with which we are so familiar, but in a mouse I think 10% of the volume of the mouse brain is white matter, whereas in humans it is close to 50%. There are many other differences, so we have to be careful how we extrapolate.

Vilhelm Bohr: I agree we have to be very careful about extrapolating studies from various organisms and mouse models to humans. We know that DNA damage and repair is an important component of the aging phenotype and that mouse DNA repair systems differ tremendously from those found in the human. Indeed there can be more similarities at some levels between Escherichia coli bacteria and humans than between mouse and man. I think pigs resemble humans more closely; rats resemble humans more than mice, so maybe we should be studying rats. Yet we do not have as many models for rat genetic knockout systems as in mice.

Sten Grillner: I accept that it is important to be cautious, but our overall experience is that when we are looking at basic features in vertebrates, then the type of receptors, ion channels, neuronal function, and the extracellular matrix are very similar. It is at the organizational level of cortical columns, etc., and the number of processing units in the cortex that there are substantial differences. Man has orders of magnitude more neurons than a mouse for cognitive processing and far more processing units or microcircuits. There are also marked increases when we go from primitive primates to man.

Aron Troen: One can think about a hierarchy of validity in looking at animal models and their relation to humans. By definition, no model truly represents reality in man, but if we try to tease apart nutrition and brain function, then animal data have a great deal of general biochemical validity, and most of what we know about biochemistry comes from animal work. One general use of animal models is in testing drugs, where preliminary animal studies are useful, but one still needs to validate the responses in humans. So, we can use the basic principles, i.e., those processes which are common across species. If transgenic mice or rats reproduce the function in the normal wild-type animal, then it is more likely to be seen in other species as well. Of course, the different dimensions of particular pathways, structural differences, and the sheer complexity of neuronal number and organization in man is important, but we have to start with the simple approach first.

Greg Cole: I use a lot of animal models and find that one can replicate some aspects of pathogenesis in one model and other features in a further model. The real issue is not so much the use of animal models but the way researchers use these models and misinterpret the results and derive poor conclusions because they do not look at the features which the animal model can represent. So, in terms of amyloid beta 42, we know from the genetics that there is a very strong case for an initiating factor. So, interventions 20 or 30 years after the initiation stage are not going to test anything relating to the initiating process. What you should do is use the animal models correctly and use a variety of animal models.

Ole Petter Ottersen: We have not really discussed the nonlinearity in the systems, and this applies to animal models. The basic physiology in the synapses between a mouse and a human is not that different, but when you perturb the system, then you amplify the pre-existing differences. For instance, the stroke model mouse and, particularly when studying these issues in gerbils, they behave very differently from humans. There is no FDA-approved glutamate receptor blocker, for instance, for use in humans because one does not get the simple extrapolated responses in humans that one expects from the rodent studies.

David Smith: Ole Petter reminded us that the white matter volume is far greater in humans than in animals, and this reflects the vast number of connections between neurons. It is not only the number of neurons but the vast number of connections. So, if there is a nonlinearity, it is probably in that area – the total number of synaptic connections. What is particularly interesting is that people have looked for ages for differences between man and animal brains, and the obvious difference is in the development of the frontal and prefrontal cortex. So, a major difference is that the rat has only a small frontal and prefrontal cortex while having the same classes of nerve cells – this is part of the evolutionary-based differences. My colleague, Peter Somogyi in Oxford, has now discovered, I think, yet another GABA [γ-aminobutyric acid]-responsive interneuron system in the rat cortex, and every time he looks in the human brain, obtained by neurosurgery, for the equivalent of a new rat subtype, he can find the same. So, basically the hardwiring is probably the same, and what may be very different are the total number of connections and the plasticity of the system in man. Humans have a phenomenal plasticity in brain function.

Steve Zeisel: Obviously, any model system is reductionist and a simplification. That is why they are useful to us, but even within the human race, we have to recognize some big differences between individuals in their processing rates of nutrients. For example 15–30% of the population have metabolic inefficiencies in the folate pathway, so we need to be careful about extrapolating even between humans.

Richard Wurtman: Species differences need careful thought. In humans, melatonin is secreted at night time, and melatonin induces sleep. However, in rats, melatonin is also secreted at night time, when they are feeding and most active! The difference is probably related to an additional GABAergic synapse inserted somewhere, so rats and humans behave differently. The second example of complexity is that if you take a rat which is hypertensive and give it tyrosine, the blood pressure comes back down to normal, whereas if the rat is in shock and you give it the same dose of tyrosine, then the tyrosine is very effective in bringing the blood pressure back up to normal. That is because neuronal responses to many nutrients depend on whether or not the neuron is firing. Responses can also vary based on differences in species, biochemical feedbacks, numbers of receptors, production of enzymes, and also on changes in physiological precursors.

ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Tom Kirkwood: We have to remember when we see declines in some parameters with age that we need to work out what is primary and what is secondary. The observation that you can have an intervention that reverses some of these changes is important as a way into establishing causality. There is now very compelling evidence that once you enter adulthood, there is no program of aging, i.e., no series of regulated changes. Changes occur in response to the altered state of cells, systems, and functions that are driven, probably at the deepest level, by the accumulation of mitochondrial DNA mutations or aggregations or simply by damage that is cumulative. So, that challenges us when we come to the intervention studies.

John Glomset: I want to talk about some work done several years ago by Han on lipids and more recently using quantitative mass spectrometry of lipids. He has used this method to examine the brains of people who have early-onset Alzheimer's disease. He measured the molecular species of lipids in the white and grey matter and looked for defects. He found in these patients specific lipids in both white and grey matter, one of which is sn-1-alkenyl-2-acyl-glycero-3-phosphorylethanolamine. It is not surprising that this ethanolamine glycerophospholipid (GPEtn) molecular species in white matter contains less unsaturated acyl constituents at the sn-2 position because the tighter packing of the more saturated fatty acids allows for compact myelin sheath formation. In contrast, the GPEtn molecular species in grey matter contains abundant polyunsaturated acyl chains, e.g., 20:4 and 22:6 fatty acids, which are well-known essential components for signal transduction. Moreover, polyunsaturated acyl chains containing GPEtn can efficiently facilitate membrane fusion between synaptic vesicles and the neuronal plasma membrane, and this fusion is an essential process associated with neurotransmitter release at synaptic terminals.

The alkenal group as an ether moiety can only be formed in peroxizomes. In the last couple of years, oligodendroglia, which of course make myelin, have been shown to have peroxizomes. Several pathways are involved, and knockout studies on these peroxizomal enzymes1 induce major disruption of myelin formation. Han has found three abnormalities in Alzheimer's disease. First, the GPEtn in the myelin in white matter is reduced in early Alzheimer disease (AD) compared with age-matched cognitively normal controls.2 Second, sulfatides, a specialized class of sphingolipids and a component of the myelin sheath, were depleted by 93% and 58% in grey and white matter, respectively, in all examined brain regions from AD subjects with very mild dementia.3 Third, there was a dramatic threefold increase in the ceramide content in the white matter of all examined regions from early AD subjects. This seems to relate to the changes in sulfatide metabolism. No alterations in other lipid classes were present at these early stages.

Han has also pointed out that the only known major genetic risk factor for late-onset AD, including both familial and sporadic cases and accounting for over 95% of total cases, is the ε4 allele of apolipoprotein E (ApoE4).4,5 Both in vitro and, more recently, in vivo data strongly suggest that the ability of ApoE to modify Aβ deposition in Alzheimer's disease may underlie the importance of ApoE4 as an Alzheimer risk factor.6 Since ApoE is a lipid transport protein, Han has hypothesized that alterations in ApoE-mediated lipid trafficking and metabolism must play a role in AD pathogenesis.7 It has been demonstrated that sulfatide is specifically associated with ApoE-containing HDL [high-density lipoprotein]-like lipoproteins in human CSF, and Han has suggested that sulfatide depletion may then occur through the turnover of the HDL particles via the LDL receptors in the brain. ApoE4 is also involved in anti-inflammatory and immune function, and the prevalence of the ApoE4 genotype is higher in tropical climates, where infectious diseases are more common. So, there may be very different evolutionary pressures.

Japanese studies have also shown that alkenyl-acyl phosphoglycerides in the ether lipids in HDL decline with age, but if you feed humans myoinositol, then they go back up to normal again, and these same changes and responses to myoinositol are seen in rat studies where comparable changes in the ether lipids in the brain and the responses to myoinositol are also observed. Inositol is available from both plant and animal sources, including wheat germ, brewer's yeast, bananas, liver, brown rice, oak flakes, nuts, unrefined molasses, vegetables, and raisins. Myoinositol is converted into phosphatidyl inositol, so this is where the control mechanism relating to ether phosphatidyl ethanolamines and phosphatidylinositol is also involved in brain signaling and is modified experimentally by alpha linolenic acid deficiency during the time of brain development.8

Greg Cole: In terms of the data from the Alzheimer brain, the biggest issue is whether you are actually losing the compartments when you see the lipid changes.

John Glomset: Studies on primate brain have shown changes in the nodes of Ranvier, which ordinarily have a distinct internode distance in a normal functioning myelin sheath. What then happens is that there is some kind of degradation going on in the myelin with repair, and in the process the internode distance is halved. This seems to slow the rate of impulse conduction tremendously.

Ole Petter Ottersen: This is a very interesting input on the composition of the membranes. We need to bring this down to the microdomain level and really understand how different membrane domains in individual cells differ with regard to the composition of the plasma membrane. So, what has been done in terms of technological advances? Will it be possible in the future to really dissect the different parts, for instance the neural membrane, and come to the exact composition of lipids? Reidun Torp has shown us, I think quite convincingly, that these amyloid plaques are formed in very close contiguity to specialized membrane domains,9 as if the composition of the membranes really matters when it comes to the precipitation of these plaques. So, what are the methodological advances here?

Raj Kalaria: If one considers the overexpressed amyloid precursor protein (APP) in mice, they show perhaps one feature, but even with the use of different strains, there are different results. So, some overall coherence has to emerge, and when we consider the effects of nutrients or feeding specific fats, we may then have to assess their effect in delaying amyloid deposition over time. However, the higher primates show that they deposit amyloid in a way that is very similar to what we see in man, both qualitatively and, in some cases, quantitatively. Furthermore, the vasculature in the higher primates is much like man, as you would expect.

John Glomset: One of the biggest challenges that lie ahead is to try and purify these membranes, and for these studies the nonhuman primate would be extremely useful. But then you need to separate the different brain cells, and this is still a challenge. We also need to remember that phosphatidyl inositol polyphosphate lipids have major roles in brain signaling and membrane trafficking, even though their concentrations might be small. There are also a number of other lipids and proteins which we need to consider.

Raj Kalaria: Axons can lose myelin and then slow down, but you do not necessarily have an abnormality in the axons if you just lose myelin. Has myoinositol been shown to affect the turnover of oligodendrocytes? This is a key question that has not, to my mind, been answered. We have started looking now in the human material, and we find in vascular cases of chronic hypoxia that the oligodendrocytes shrink and some disappear; we do not seem to see turnover or replication in these cells.

John Glomset: Studies recently in the United States suggested that peroxisomes provide oligodendrocytes with an essential neuroprotective function against axon degeneration and neuroinflammation, but this is not linked to glial survival necessarily.10,11

Raj Kalaria: Neurogenesis is interesting. It is has recently come to light that there are precursor neural stem cells in the subvesicular zone as well the dentate gyrus.

Steve Zeisel: The problem is that the neural precursor cells in the superventricular zone are probably one-tenth of one percent of the total cells in the brain, and so even if they quadruple the number, it would not change brain volume appreciably. But if you argued that other stem cells are just as susceptible, then this could be a more general phenomenon.

VASCULAR ISSUES

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Philip James: We heard from Raj Kalaria about both macrovascular and microvascular problems. Raj was emphasizing that microvascular disease could lead to an Alzheimer's type syndrome or dementia with multiple damage, whereas Marshal Folstein was emphasizing the maintenance of cognitive function, which may or may not relate to microvascular disease. Are we talking about anything different?

Bo Angelin: We have to realize that aging is one thing, and age-related disease is another. We tend to talk about a change in a number of risk factors for age-related disease which will obviously affect lifespan. The most thought-provoking idea comes from experiments with Caenorhabditis elegans, where you see the importance of stochastic issues: when diet, environment, and genes are all identical, you still find a stochastic distribution of effects, which means that there must be something else influencing this. These results amplify the basis for the theory of damage. Genetics will influence the susceptibility to damage in an individual, but strictly speaking, aging is an independent feature.

Raj Kalaria: When we look at the aging literature, there are numerous studies that show a 10–20% effect of some interventions in people; these tests are undertaken at several times, e.g., in middle age as well in the elderly. However, it is the additive effect of multiple interventions over a long time scale that is important. When we look at specific cell types either in a particular part of the brain or elsewhere in the body, we may see the responses to interventions, but in aging, as we assess the loss of some cognitive measure, we are really looking at the additive effects. Only when there is a clear event such as a stroke do we see an immediate impact, but atherosclerosis of both carotid arteries with a lowering of the blood flow with age produces a marked cumulative effect without a specific clear event. Then I think the same reasoning applies to spine density – we could count spines forever, but it is really the cumulative effect with perhaps very specific regional changes which is evident in functional terms. If, for example, the hippocampus is affected by aging, then it will impact on other parts of the brain, and the same is true for the frontal lobe: if there are white matter lesions early on, then the loss of executive function is an early part of this dementing process and very common in the general population. Given the connectivity of the brain, small changes in specific regions of the brain, e.g., the frontal lobe, lead to an additive general effect on cognitive function, and this may be much more apparent with supposedly regional changes and does not require the extensive damage that is observed in people with general dementia.

Marshal Folstein: We have to separate the brain from the “mind.” The word “dementia” refers strictly to the “mind” as a collection of cognitive symptoms; it is a deterioration of multiple, cognitive functions and a clear consciousness. The dementia syndrome is the product of several different discrete pathological processes, one of which is Alzheimer's disease. Alzheimer's disease is recognized by a particular change in the brain, and we know, because of Down's syndrome and other syndromes, that these changes can sometimes be caused by specific genetic mechanisms. Some cases might be associated with vascular disease, so let us separate them out and what we have left is, first, a population that has some subclinical cases relating to vascular disease. Then, when we look at a population with dementia excluded, we see the conventional things, i.e., that aging and years of life are associated with atrophy of the brain. Brain atrophy can be associated with the number of small vascular infarcts, but the other issue is the reduction in the volume of what is called “white matter hyperintensity.” There is some argument as to what the vascular component contributes to this hyperintensity – I think most people agree that at least a significant proportion of it is due to ischemic lesions in the white matter. So, my measure of age-associated brain atrophy has a component of vascular atrophy. When I looked at vascular-associated brain atrophy, it was associated with the classical decline in memory and in central processing speed that we generally think of as cognitive aging. Now, is it possible that those same factors will eventually turn into or precipitate Alzheimer's disease? Yes, some of them might, but I always want to look at the neurons and the glia and not just the vascular problem. I now think that there is a vascular component that is associated with cognitive deterioration, and it is different from the processes underlying Alzheimer's disease, which has a different etiology.

Bo Angelin: We know we all have some atherosclerosis, but the effects may be small and relating simply to the reactivity of the blood vessels. We could say that without the circulating LDL cholesterol, there would be no atherosclerosis, but the problem is that you probably need a small amount of LDL, although HDL can actually do a lot of the essential lipid transfer. Rats and mice have LDL levels that are at least fivefold lower than ours, and they obviously survive well from a vascular point of view. People say that dementia develops in those with a normal cholesterol, but I would say that very few humans have an optimum cholesterol level, and what Westerners call a normal population cholesterol value is high. So, I think the vascular component of dementia is probably related to atherosclerosis and should be seen as a disease process. Then you have some of these genetic diseases, which affect arteriogenesis, etc., but that is something else. So many of the positive effects on mental function are actually related to subclinical atherosclerosis, e.g., HDL changes and the effects of physical training, etc.

Ricardo Uauy: HDL lipid levels are protective, and ApoE has been known to be the carrier of DHA into the brain, so we have at least two important transport systems. In HDL terms, we normally think of reverse cholesterol transfer from the vessels into the liver by HDL, whereas ApoE is involved in transferring essential fatty acids widely.

Marshal Folstein: I agree vascular sensitivity is important. When microalbuminuria occurs in mild atherosclerosis in the kidney, one finds significantly associated white matter hyperintensity in the brain. So, there is some very close relationship between relatively mild vascular disease in the kidney and the brain. Whether the kidney is mediating other effects and not simply reflecting the general atherosclerosis, we do not know.

Ingmar Skoog: Even what we call normal cognitive decline, if there is such a thing, relates to vascular factors and almost every vascular risk factor you could think of.

Irv Rosenberg: Raj Kalaria emphasizes that there may be not only size differences but also pathogenic differences between macrovascular and microvascular changes. We have also found microvascular changes in the brain and microalbuminuria,12 but I think it is important to at least hold open the possibility that some of the microvascular changes that we are seeing in the brain, although sharing some of the same risk factors with vascular disease, may not be the same process.

Philip James: We heard earlier that high blood pressure in middle age is a very bad predictor for maintaining your mental agility, so have studies been done on cerebral atrophy in those populations where you have very low blood pressure and very low blood cholesterol levels and no obesity? Do these populations have minimum brain atrophy as they get older?

Ingmar Skoog: I do not know of these studies, but it is evident that brain atrophy, Alzheimer's changes in the brain, clinical Alzheimer's disease, and vascular dementia are all related to the traditional vascular risk factors. There are a number of studies – I think the Framingham Study is one of them13– where we actually found that hypertension or high cholesterol or diabetes mellitus increases the risk of brain atrophy in old age, independent of whether you are demented or not. But dementia is also related to these factors.

David Smith: I am glad to have Ingmar's authoritative answer on that. We found that if you stage the Alzheimer's pathology according to the Braak staging system,14 the distribution of the neurofibril tangles started in people in their late twenties, and at that stage there are very few tangles localized to the entorhinal cortex. But, gradually, as they looked at 2,000 random autopsies and the distribution of neurofibril tangles over the age range, they developed the hypothesis that these tangles spread from the entorhinal cortex into the hippocampus and gradually throughout the neocortex. They then came from Frankfurt to Oxford and looked at our brains and did the same staging on people where we had cognitive information. We were then able to show that at certain stages in this process of tangle spreading, people do begin to show cognitive impairment and then later dementia. So, you can use this now as a marker, i.e., the presence of neurofibril tangles in different parts of the allocortex and the neocortex; these tangles are a marker of a stage in the development of Alzheimer's disease over a period of 40 or 50 years. We found that this development is correlated with cognitive test scores, but people who have small vessel disease have a lower cognitive score for the same tangle stage.15,16 So, it was almost as if the vascular problem was adding to, or synergistic with, the tangle pathology. So, they had, in clinical terms, Alzheimer's disease, but you would not diagnose it as such if you just looked at their brains; they had pathologically early Alzheimer's, but the vascular problems were bringing forward the cognitive expression of Alzheimer's disease.

Irv Rosenberg: I am comfortable with Laura Fratiglioni's use of terminology where “dementia” is an essentially advanced stage of cognitive dissolution rather than using the word “Alzheimer,” which took over from “dementia” in the 1970s and 1980s, but there are reasons for retaining some specificity in the use of these terms. There is now ample evidence that vascular factors are involved in brain aging and significant cognitive decline. On the matter of how blood vessels react in brain to such things as exercise, I think we have to recognize that not only do we have to be thinking about the plasticity of our ability to create blood vessels and angiogenesis in response to stress, but there is also an exquisite regulatory system whereby these blood vessels respond to systemic changes in blood pressure or stress factors. So, it is not just a question of synthesizing new blood vessels or increasing or decreasing the capillary bed needed for the delivery of oxygen and other nutrients to the brain, but there is a minute-to-minute regulation of the function of those blood vessels that has a lot to do with the systemic blood pressure and other systemic factors.

Steve Zeisel: So, let me suggest a hypothesis relating to brain atrophy. Early-life exposures can change not only the rates of neurogenesis and numbers of neural precursor cells but also the numbers and rates of proliferation of endothelial progenitor cells. So, when there are multiple vascular accidents, the response is to then stimulate stem cells or, rather, the progenitor cells that are semi-differentiated, to then try to repair the damage and recreate new cells. A decade ago, we suddenly found that there are neural precursor cells that can regenerate, so perhaps aging is just the process of running out of these seeded progenitor cells that are capable of reproducing and becoming new nerve cells and new endothelial cells. This limitation is less likely for the endothelium and more probable for neurogenesis, since we know that the neurogenesis rates drop a hundredfold from early life to middle age. Maybe we are running out of stem cells, and that is why we are running out of brain.

Marshal Folstein: I am not so sure about this, because the idea of a stem cell reserve implies one is somehow protected from the onset of dementia. In fact, our data suggest that people with a higher developmental potential – that is, the higher verbal intelligence – oddly show more atrophy once the process has started. So, on the one hand, a good brain capacity seems to be protective, but once a pathological process has started, you deteriorate more quickly, which I do not understand.

Patrick Stover: This concept of regenerative capacity was highlighted in a recent paper17 which showed that the number of muscle stem cells in an aging mouse did not change, but the differentiation potential did. When they did a parabiosis study with the young mouse, it actually restored the differentiation potential in both muscle and liver to the older linked mouse. So, there are also circulating factors that determine whether or not the stem cells are going to be able to become functional cells, and this could also involve the brain.

NUTRITIONAL ASPECTS OF PLASTICITY

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Aron Troen: I would like to consider the issue of brain malleability and other compartments in the brain, specifically the vascular compartment and its plasticity in the adult. We need to consider the plasticity and repair mechanisms of the adult brain. We have data where we have considered the epidemiological literature looking at associations of brain abnormalities with vitamin B deficiencies and used animal models – in our case young adult animals – to assess the impact of different perturbations of one-carbon pathways. These studies attempted to dissect apart the mechanisms which relate to specific outcomes using biochemical and behavioral outcomes as well as anatomical changes. We have used both apolipoprotein E deficient mice and wild-type C57 mice which are B-vitamin deficient in folate, B12, and B6 and find no evidence of neurodegeneration or demyelination but a reproducible and robust behavioral deficit in the water maze test.18 We then looked at the microvascular compartment by doing stereological counting of the total capillary length in the hippocampus and found a 30% reduction within 10 weeks in a weanling animal that correlates very strongly with the escape latency from the water maze. So, animals that have a larger vascular hippocampal length perform much better; they escape more quickly from the water maze than animals that have shorter total capillary length. However, if we give other mice a selectively high methionine diet which produces only a moderate elevation of homocysteine, then this induces a severe 50% reduction in the number of microglia in the same brain region, but without the behavioral deficits.19 So there are differential effects on different compartments under similar metabolic conditions. Changes in microglial proliferation or survival might, of course, have other types of toxic or adverse effects. Data from Inna Kruman20 in Mark Mattson's laboratory have shown a reduction in neuronal proliferation induced by folate deficiency, with a reduction in glial numbers and capillaries, the effect being either an inhibition of proliferation or a toxic effect. She finds that the abnormalities of one-carbon metabolism repress the proliferation of cultured embryonic multipotent neuroepithelial progenitor cells and affect cell cycle distribution, thereby suggesting that dietary folate deficiency inhibits proliferation of neuronal progenitor cells in the adult brain and therefore neurogenesis. The capillaries produce trophic factors as well as the supply of metabolites and their substrates to the brain. These trophic factors act locally at the neurovascular junction to maintain neurons both during development and subsequently in the adult. New work is now looking at the molecular cross-talk and signaling pathways and, for example, neuropilin-1 vascular endothelial growth factor A (VEGF-A) interactions.

Patrick Stover: In terms of microvascular plasticity, there is wonderful work coming out of Joe LaManna's lab at Case Western Reserve [University],21 where they show that exposure to certain metabolic stresses, e.g., hypoxia in adult rats, induces a doubling in the capillary density in the adult rat cortex within 3 weeks. There are also responses following an insult such as stroke. Ketogenic diets as well as hypoxia will induce a robust upregulation of capillaries, actually in muscles as well as the brain. These same changes take place after a stroke, with an upregulation of angiogenesis and neurogenesis. In the penumbra around the stroke, folate deficiency increases the lesion's size, and genetic models that are deficient in downstream aspects of folate metabolism such as uracil DNA glycosylase deficient mice have a larger penumbrium. One can then look at the DNA damage that relates to folate deficiency specifically and its relationship with the extent of the lesion. Finally, folate deficiency induces choline deficiency, and choline deficiency induces folate deficiency, so I think there is a great deal of cross-talk between these pathways.

Richard Wurtman: You mention that a ketogenic diet increases capillary formation, and ketogenic diets have become quite popular as a treatment for epilepsy.22 I wonder whether this is another stress-induced response.

Aron Troen: Michelle Puchowitz23 in Joe LaManna's lab has been exploring that and found a marked increase in the glucose transporter GLUT-1 as well as the monocarboxylate transporter as a part of what is considered an increased vascular density but without any change in blood flow, but in humans ketosis also increases the blood flow, and this seems to facilitate transport. Puchowicz has also shown neuroprotection in diet-induced ketotic rat brain after focal ischemia,24 so the speculation is that the same metabolic switches, including hypoxia-inducible factor-1α, a transcription factor that mediates adaptive responses to hypoxia, is involved.

Bo Angelin: Perhaps one of the things that is stimulated by a ketogenic diet and metabolic stress is the selection of fibroblast growth factor 21, which is a metabolic regulator in the adult and obviously of critical importance in early neural development. Unfortunately, you cannot measure it in mouse plasma.

TELOMERES

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Philip James: Tom Kirkwood in his discussion emphasizes the accumulation of multiple forms of damage as a major feature of so-called aging. So where do all the stories about telomeres come into the picture?

Vilhelm Bohr: Tom Kirkwood's overview recognized there are many theories of aging, and while he dismisses the genetic regulation of aging, there still is a lot of work suggesting that there are certain genes which regulate the aging process, even if the genetic regulation has a smaller impact than environmental factors. For example, studies in twin populations suggest that maybe 25% of the age-related phenotypes are of genetic origin. The damage and repair theories are actually well set out, with specific ideas about the nature of the damage and the types of DNA repair. The telomeric hypothesis could be in part genetically regulated, dealing with the function of the telomeres – repetitive sequences at the end of chromosomes – which protect the chromosomes from genomic instability. Telomeres shorten with age due to a simple process involving replication at the end of the chromosomes. This has been shown extensively in fibroblasts and various tissue culture settings and then relates, for example, to Hayflick's theory of aging,25 which is that senescence develops as the number of replications progress, so a cell has a limited number of cell divisions. But that is a tissue culture phenomenon, and we do not know if the telomeric shortening also happens in vivo and in tissues in humans.

Philip James: Are you implying you cannot measure telomeres in vivo?

Vilhelm Bohr: A lot of measurements have been done, but it is not so clear what happens in various in vivo systems rather than in cell cultures.

Philip James: I understood that Nick Hales26 had studies published in Nature where he calorie or protein restricted pregnant rats and showed that these animals with caloric restriction have a longer lifespan and their telomeres were longer, He also showed, however, that brain telomeres did not show the gender- and age-related differences in length seen in other organs.27

Vilhelm Bohr: There are endless studies of telomeric length in different tissues and under different treatments, but the in vitro results are not necessarily duplicated in vivo. Furthermore, average telomeric lengths are usually taken, but we know that telomeres differ in length depending upon the chromosome. So, some chromosomes, the fourth, for example, have a shorter telomere. So, now assays are being devised which test the length of individual chromosomes rather than the overall average. Yet the signaling process involving the telomere is extremely important and relates to another major theory of aging, i.e., mitochondrial degeneration. Mitochondrial DNA damage is much more prominent in aging than genomic DNA damage. DNA repair processes are poorly understood in mitochondria, but links are beginning to emerge between telomere signaling and mitochondrial function. There are several proteins as well as an RNA telomerase, the enzyme that lengthens the telomeres, involved in the control of telomeric length and of signaling in mitochondria and genomic DNA. A number of telomeric proteins called shelterin proteins,28 the six-subunit protein complex that protects chromosome ends from the DNA damage, regulate telomere length maintenance by telomerase. Shortened telomeres often display extensive DNA processes, including strand breaks in the telomeres. So, again there are lots of very important meaningful correlations, but we should probably look at specific chromosomes for specific functions in relation to aging. When the telomeres shorten, some proteins – perhaps P16, P53, or other proteins – might be signaling at the cell cycle regulation checkpoint. Therefore, several different chromosomal controls may be involved in aging,29 or it could be just the shortening of one telomere that signals the whole thing.

Jacob Selhub: We have just published analyses relating telomere length and folate status30 and the genetic relationships to plasma homocysteine, plasma vitamin B12, and plasma PLP [pyridoxal phosphate].31 The results of our study showed that the plasma concentration of folate was associated with the telomeric length of blood mononuclear cells in a nonlinear manner. When plasma folate concentration was above the median, there was a positive relationship between folate and telomere length but an inverse relationship between folate and telomere length when plasma folate concentration was below the median. The MTHFR 677C>T polymorphism was weakly associated with increased telomere length at below-median folate status. So, we think that folate status influences telomere length by affecting DNA integrity; a low folate status means more uracil incorporation into DNA and less protection of the telomere.32 Then, there can be epigenetic regulation of telomere length through DNA methylation.

Ricardo Uauy: I think that the telomeric idea and Hayflick's hypothesis are fine when we consider dividing tissues, but aging, especially brain aging, is affecting postmitotic cells. Nonmitotic cells are driving the aging process in the brain, and the question then is one of the inter-relationship between changing brain function and the control of metabolism generally.

Vilhelm Bohr: The telomere is a good marker for some things, but there are also many instances where it does not fit with the relevant endpoint. So, in meetings on aging and telomeres, there is greater emphasis on the mechanism of telomeric maintenance and whether the DNA repair processes are more or less effective in the telomere and the ability to respond to double-strand break signals. There are some very interesting examples that go against the oxidative theory of aging. For example, there are some birds that live a very long lifespan, maybe 90 years or so, e.g., parrots. They have a high metabolic rate and generate enormous amounts of oxidative damage in their DNA and other macromolecules, but they do not seem to have a very efficient DNA repair process. So how do we explain that?

Steve Zeisel: Even if some birds do not comply with general findings, there seems to be a very strong association between metabolic rate and lifespan. So, when we are trying to consider genetic influences, it could well be that the genetic control of metabolism affects the damage done by high metabolic rates – whether this damage is mediated by oxidative changes and by other factors involved perhaps in changing the relationship between life span and metabolic rate.

Cutberto Garza: Is telomeric shortening relentlessly driven by intrinsic or extrinsic processes or a combination of the two?

Vilhelm Bohr: A combination, but the major feature is an intrinsic so-called end-replication process, which means that you cannot continue to replicate endlessly. So, telomeric shortening is like a clock of cellular senescence. In addition, you can also have damage to the telomere. So, there are gene-rich areas where base modifications by oxidative lesions are frequently formed. These are hot spots for oxidative damage in the telomere as well as the constant shortening relating to the end replication process.

Cutberto Garza: This then is in keeping with Ricardo's comment that there are mechanisms for a postmitotic cell to be affected; telomeric shortening reduces the cell's flexibility in responding to future stresses.

Irv Rosenberg: Do we see different responses in different organs in telomeric length terms?

Vilhelm Bohr: There are significant differences between tissues.

OTHER INFLUENCES: STATINS

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Bo Angelin: There is a provocative paper from the West of Scotland Study, which showed in the placebo group that there was a relationship between survival and the telomere length in blood leukocytes. Individuals in the middle and the lowest tertiles of telomere length had nearly double the risk of developing a coronary heart disease event than those in the highest tertile. Statins reduced the risk of coronary disease in those with shorter telomeres.33 Then the issue is how to restrict calorie intake in humans; actually, one can have intermittent fasting, e.g., on alternate days, and this will then induce different mechanisms.

Kay-Tee Khaw: Like Bo Angelin, we should be highlighting clinical studies. Cardiologists have shown that the diseased heart has a shorter telomere length, and it is surprising that a crude measure of peripheral lymphocyte telomere length is actually a very strong predictor of mortality. Whether it is just a marker of the general aging process or whether it is on the causal pathway, we do not know. But we do know that a lot of lifestyle factors are associated with a shortened telomere length, e.g., smoking.34

Richard Wurtman: How could telomeric shortening influence brain function?

Greg Cole: There is minimal evidence for telomere shortening with clonal exhaustion and replicative senescence in the brain: the astrocytes and microglia do not get anywhere near their proliferative limit from evidence that I have seen, but there is mitochondrial damage, and that relates to the insulin signaling problem, IGF1 [insulin-like growth factor 1], and neurotrophic pathways. Treatments with insulin-related drugs such as the PPAR [peroxisome proliferator-activated receptor] gamma ligands and glitazones are being tried in an effort to improve the energy deficits in the aging brain. Nutritional interventions like DHA also impact the PPAR-gamma signaling system. It is one of the ligands which could regulate the replication of mitochondria; out in the dendritic fields of the brain, you have a problem with the protection of mitochondria.

A LIFESPAN APPROACH TO AGING

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Ingmar Skoog: An epidemiological comment. One of the most constant findings in dementia is that head size is related to dementia. Head size is, as you know, a measure of premorbid brain size and the size of the brain at the age of 5 years. So, I think this could be in keeping with early programming of brain development.

Aron Troen: I agree we do need to distinguish between early development and later life. If we want to understand the effects in later life and consider early nutrient-dependent imprinting, then I think the choline deficiency model is interesting. If there is a selective limitation of choline supply in early life, this induces persistent effects in the adult despite subsequent free access to choline. Some of the choline-processing enzymes are imprinted and downregulated or perhaps expressed at lower levels in those animals that were deficient. So, we may be looking at the effects in response to diet in old age, but the organization of the responses may be set up very early and are not then as amenable to modulation.

Ricardo Uauy: I would question the idea that aging is dependent on evolutionary pressures: after the time of reproduction, how can the evolutionary pressures have much impact? So, basically, it is a man-made environment that will be defining successful aging: only very recently has man had a chance to age beyond the reproductive span.

Jon Storm-Mathisen: It is not so simple, because humans are a social species, and old people may affect the success of their offspring.

Kay-Tee Khaw: I was very struck by the evidence Ole Petter showed about plasticity. We heard how the microenvironment, such as nutrients, may affect the brain's plasticity, but I would like to hear a bit more about sensory input, because there is such a lot of evidence, for example, from nerve damage studies, that if you try and repair nerves, then the functional regain depends not only on giving components that will help nerve repair, but also critically on the sensory input at the time of retraining. So, is there a critical point at which the micronutrient might have to interact with sensory input? There is a lot of evidence about whether children learn better before or after meals, for example, so I just wondered whether there is something related to that?

Ricardo Uauy: I think that nutrition interventions clearly need to separate how much we are doing introducing the concept of age-related disease. In the animal data, this is very confounded because the animals die not because of old age but because of tumors, mainly. One can prevent tumors by manipulating the diet, and the same thing seems to apply to humans: if you take people with BMIs around 21, in general they will live longer than people with BMIs over 30. So, until you separate these issues out, it is difficult to look at aging and lifespan per se.

Ole Petter Ottersen: When we are talking about plasticity, we need to consider whether we have the same kind of plasticity in older individuals as in younger individuals. For instance, if you challenge old people with a paired cue recall test, then young and old people tackle these tests in a very different manner. Young people tend to involve just one hemisphere, whereas older people might easily process this challenge with the use of both hemispheres. So, plasticity is something that extends far beyond individual synapses.

Marshal Folstein: I have just two simple points to make. What happens from microsecond to microsecond at the synapse is undoubtedly important for some processes, but the wiring together of neurons as they fire, which some people take as the psychological model of plasticity, happens over a much longer time course. So, the example of vocabulary changes in aging, with a functional increase rather than a decrease with age, is an example of an entirely different time course. So, we need to remember that these various biochemical processes and functions operate in different time dimensions, and probably the effect is specific to a time dimension. There is also no question that a damaged brain, whether because of age or a stroke, processes information differently from a normal brain. So, there is no question that old people's brains work differently: the question is how and why they work differently. The different aging processes can be a feature of growth and development as well as degeneration and repair.

Philip James: I had the impression that a 25-year-old might process in a very different way from a 75-year-old.

Marshal Folstein: Probably in some areas, but the one thing we can be absolutely certain of, based on decades of research, is that the 75-year-old will be a lot slower than the 25-year-old. Whether he would be more accurate or have more capacity or a different capacity probably depends on the nature of the task.

Raj Kalaria: If one tests middle-aged people after a stroke with very simple motor tasks, some of these individuals' responses might reflect the value of brain reserve. In our post-stroke cognitive impairment dementia study, after the first stage, where all were initially cognitively intact, we then found on testing over a period of time that a third of them clearly improve in neuropsychometric measures, but others do not. So, in this strictly comparable group, the marked differences in subsequent response may well imply differences in plasticity and perhaps brain reserve which relate to features other than simple vessel injury, which we assessed by looking at blood flow in the brain with MRI [magnetic resonance imaging] methods. So, the issue of plasticity may be important even in the elderly.

Cutberto Garza: So, it is not basic neuronal or synaptic processes that are subject to senescence, but the architecture or the organizational structure in which these processes occur?

Tom Kirkwood: I think it is fairly clear from all kinds of evidence that it is likely to be both. You will have changes that affect the connectivity, the networking structure; you will also have changes that will affect intracellular molecular processing. I think obviously one of the challenging questions is to understand where you see functional change occurring and which of those changes may be dominant. I suspect it will be quite hard to disentangle.

Cutberto Garza: What is the basic evidence that neuronal synaptic processes age? Not the three neurons working together, not the networks, but the basic neuronal synaptic process itself?

Richard Wurtman: The synaptic number is also important, with the number allowing signal amplification.

Richard Black: I thought that there were some data showing that as the cell ages, the amount of neurotransmitter released from a particular bud diminishes and that you then get compensation from other adjacent buds that try to maintain the overall level. So, there is actually increased function at some of those gaps.

Richard Wurtman: In all clinical trials for Alzheimer's disease, there is a well-recognized international problem; namely, that people are not getting worse as fast as they used to, and nobody knows why. It may be related to regional differences in fish oil or something else, but it is a big problem in doing clinical trials. In one of our clinical trials, there was no reduction with age in the scores of the placebo-treated people, which is not really surprising. In clinical trial terms, one needs to recognize that, experimentally, DHA by itself will stimulate dendritic spine formation and synaptogenesis, but it is much less effective than the combination of DHA and uridine and choline. In our completed study, those with very mild Alzheimer's disease showed an improvement in cognition after nutritional intervention, which was very significant (P<0.01 or P<0.001), and the benefit was greater in people with just mild mental impairment as defined by Marshal's criteria.35 Our current study of 500 subjects under way in the United States will have very mild and moderately affected people who will receive DHA, uridine, and choline. We do not know whether it is better starting to treat early on to suppress the early phase of disease progression or to wait until there is something substantial to see when assessing whether we can reverse the problem.

Marshal Folstein: It is extraordinarily important to take into account either measures of verbal intelligence, like the adult reading test, or a factor that is closely related but confounded with it, which is schooling. This influences two things: it influences the susceptibility to the onset of Alzheimer's disease, and, secondly, schooling and intelligence are highly intercorrelated, with a variety of the cognitive outcomes. For example, when you look for homocysteine effects on cognition, if you control for intelligence first, you get completely different answers.

Philip James: What is the complication of schooling? You are brilliant because you went to school, or because you went to school, your brain is much better?

Marshal Folstein: They are related but independent: there clearly are factors that are influencing your verbal intelligence that are independent from schooling, but schooling clearly also helps. Some people use intelligent school as a single concept.

Thomas Cederholme: In our study,36 it was actually verbal fluency that was improved most during the first 6 months of our study. I think we have to be very cautious using the randomized controlled trial as the truth when we deal with such slowly progressing disorders. We also need to rely on good, prospective, epidemiological studies before we draw our conclusions as to whether a nutrient, e.g., DHA, is good or not.

Ricardo Uauy: The other major problem is in the use of depression scores. If you do not adjust for depression, you will find a lot of spurious results. Furthermore, when we relate effects to foods rather than specific nutrients or compounds, we also now need to take account of genetic heterogeneity. So, with DHA, there is genetic heterogeneity in the conversion of linolenic acid to DHA, so when you correlate the results of mental tests with plasma levels of DHA, they are much stronger than the relationship to dietary factors. So, I think we need to look at all the possible approaches.

LIPIDS AND AGING

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Patrick Stover: What is known about the lipid composition of membranes as one ages? We saw some anatomical changes in the brain, but what are the associated changes in the lipid make-up of the brain – are there any with aging?

Ricardo Uauy: We used to say that brain lipids could only be changed during development, but now that has been challenged, and in fact we have dietary modulation of the membrane composition of the adult brain demonstrable at least in mice. We have to accept that there is slow membrane turnover, and in the case of the retinol photoreceptors, for example, it is very clear that they can be modulated by diet at various stages of life even though the turnover is very slow. In the case of lymphocytes and the liver, there are much more rapid changes, and obviously these changes are reflected in functional terms. We should also be thinking about lipid rafts and not just a homogeneous lipid layer in the membranes, and the rapidity of change in composition and its dynamic relationship to memory. Lipid-protein interactions are also defining transport and receptor activity, so it is a much more interactive model that we need to consider. There is not only overall fluidity but also lateral mobility and all sorts of fluidity states at the micro level that have tremendous functional implications.

Kay-Tee Khaw: Is a raised HDL cholesterol level an indicator of a potentially greater carriage of lipids across the blood-brain barrier, and does that HDL system influence the fat in brain myelin?

John Glomset: Obviously, people have been thinking about HDL as a carrier of cholesterol in reverse cholesterol transport to the liver, but I am sure they have not been thinking very much about HDL as a positive carrier elsewhere. Ether lipids and alkenal acyl phospholipids are in HDL and can be transmitted from HDL into cells. It has been shown that the content of this particular kind of lipid also increases in erythrocyte membranes, and these lipids have phosphoethanolamine components, so if they were to be transferred by exchange reactions between the surfaces of lipoproteins and membranes, for example, there is an enzyme, translocase, an ATP-dependent translocase, that takes ether lipids and flips them across from the outer to the inner membrane leaf.

Philip James: Since several lipoproteins are carriers of these important ether-related fatty acids and phospholipids, is this the fundamental linkage between lipid metabolism in the liver and brain uptake of lipids?

John Glomset: What happens to HDL in the brain is unclear, but HDL takes lipids from the liver, and there is a phospholipid exchange protein in plasma, so that VLDL [very low-density lipoprotein], for example, through the exchange protein, has lipids transferred from the HDL fraction.

Richard Wurtman: What does the brain get from the blood that it needs in order to synthesize these phospholipids? You can show that individual nerve terminals contain all the enzymatic machinery that is needed to synthesize all the phosphatides, for example. So, there is no need for the circulating phosphatides to get into the brain. I do not know about the ether-linked lipids, but there are very few compounds which have been demonstrated to travel from blood into brain in relation to lipid metabolism.

Ricardo Uauy: The brain does not make the essential lipids, e.g., DHA.

Richard Wurtman: Agreed, the only lipids that the brain cannot make are the essential fatty acids. I know because we have studied the issue.37 The components needed for phosphatidylcholine synthesis – the fatty acids, etc. – all transfer across the endothelial cells into astroglia, and once synthesized in the brain, they are exchanged from cell to cell.

John Glomset: There is an exchange between plasma phosphatidylcholine and the outer leaflet of cell membranes, but as soon as you have ethanolamine in there, then the ethanolamine gets trapped inside the inner leaflet, and that is a whole different story.

Richard Wurtman: It is not generally accepted that circulating phosphatidylcholine gets into the brain. Synaptic zones are completely capable of synthesizing phosphatidylcholine, and by methylating phosphatidyl ethanolamine, free choline is generated in the process.

NUTRITIONAL INFLUENCES

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Cutberto Garza: A series of breakfast studies done 30 years ago looked at learning after fasting or after meals and found improved learning and other cognitive functions after breakfast as opposed to fasting.38 Our testing now would be far more specific, but we need to remember these studies were always being done in the young. I am not aware of any similar studies in the elderly, and I do not know if one would see the same response.

David Smith: When we think of the impact of nutrients on brain function in adults, we must not just focus only on micronutrients; it is well established that simply giving adults glucose improves their memory performance, so there is an energy supply/metabolism issue as well. However, we also have to recognize that with micronutrients, transport proteins and enzymes have a low affinity for the micronutrient. For example, in the folate system, a large number of the relevant enzymes are actually of low affinity, so by increasing the level of folate input, you are going to increase the throughput of folate in these systems.

Jon Storm-Mathisen: When you start to modify nutrient intake and change metabolic precursors, then the body might respond by adjusting the levels of receptors and their activity as well as all the protein/enzymic responses, so this is also something that has to be taken into account.

Richard Wurtman: We have tried to ask over the years whether, in making more of a precursor available for the transmitter, there are any functional consequences that might help the animal. The best examples are tryptophan and serotonin. Under normal conditions, depending really on the composition of breakfast, serotonin synthesis and release either go up or go down. If you have some orange juice, coffee, and sugar, you are going to make more serotonin; if you have bacon and eggs, you are not. And then you can speculate that maybe this is a signal to the brain to decide whether to eat and what to eat, and maybe it has an effect also depending on the time of day. The effect is probably greater in women, in part because the brains of women, I am sorry to say, only make about two-thirds as much serotonin as men's brains. Maybe this has evolutionary significance so that women with less serotonin will hear the baby cry. In the case of the other neurotransmitters, we can only speculate, but I don't think there is really a strong argument about why acetyl choline production is dependent on choline intake.

Aron Troen: Another example of differential effects from alterations in the one-carbon pool metabolism is the effect of folate deficiency in reducing dopamine release in the nucleus accumbens. We are finding that adding methionine does not restore dopamine release. So, we have to be very specific in the outcomes that we measure in these experimental models.

Ole Petter Ottersen: I miss one very important aspect when it comes to Dr Wurtman's and Dr Smith's comments about unsaturated enzyme systems, and it relates to Tom Kirkwood's evolutionary approach. If enzymes are set at a level where they are not saturated, could it be because there has to be a trade-off between the beneficial effect of reducing their affinity and the negative effects of having high-affinity enzymes? I think we must take into account evolution when we discuss these matters.

DHA, DEVELOPMENT, AND BRAIN FUNCTION

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Christian Drevon: I want to make a few points concerning the fatty acid composition of the brain. Sometimes one gets the impression that DHA is the only important fatty acid in the brain, but that is not the case at all. If one assesses the fatty acid pattern of human brain and takes, for example, the prefrontal cortex, then saturated fatty acids dominate (36%), particularly palmitic and stearic acid; the next abundant is omega-9 fatty acid oleic acid at ≈30%; the omega-3 fatty acid DHA is at ≈20%, and the omega-6 arachidonic acid at ≈10%, but the omega-3 fatty acids linolenic acid and eicosapentaenoic acid (EPA) comprise <1%39 and the omega-6 fatty acids linoleic acid and docosapentaenoic acid ≈ 2% of total brain fatty acid composition. So, we tested their potential importance by supplementing women from 18 weeks of pregnancy with 10 mL of cod liver oil or 10 mL of corn oil, with the same amount of fat-soluble vitamins in the oils,40 and continued supplementation throughout pregnancy and for 3 months after delivery. Then, using the Kaufman Assessment Battery for Children at 7 years of age, we could not find any growth or statistically significant increases in cognitive processing in the offspring. However, there was a clear tendency to increase cognitive scores in the cod-liver-oil-supplemented children at 4 years, and when the plasma lipids of the pregnant mothers and children were tested, the omega-3 very-long-chain PUFAs [polyunsaturated fatty acids] in maternal blood at 35 weeks of pregnancy as well as in neonatal samples positively correlated with scores on the sequential processing scale in the children at 7 years of age. Later, critics suggested that, given the normal large size of babies in Norway and women's health, we were minimizing the chances of picking up a supplement effect. Then, we did another study on premature infants born with a body weight below 1.5 kg.41 These infants lacked the possibility of gaining all the important nutrients via the placenta for, on average, the last 3 months of their fetal life. We tested the brain's electrophysiology by EEG using six different recordings, together with visual challenges. With these tests we then assessed the impact of using human milk supplemented with 32 mg of DHA and 31 mg of arachidonic acid per 100 mL of human milk, starting 1 week after birth and lasting until discharge from the hospital (after on average, 9 weeks). Cognitive development was evaluated at 6 months of age by using the Ages and Stages Questionnaire and event-related potentials, a measure of brain function related to recognition memory. The data show that supplementation with DHA and arachidonic acid for very preterm infants fed human milk in the early neonatal period was associated with better recognition memory and higher problem-solving scores at 6 months.

Ricardo Uauy: These data extend what has been available for the last 10 years. There are now quite long-term follow-up studies showing the effects are more marked with early interventions. There are demonstrable electrophysiological changes and behavioral changes in studies of DHA supplementation of preterm and term infants. These tests include visual acuity, processing potentials, electron tomography, and visual recognition. Why vision? Because DHA is highly concentrated in retinal tissues and also in occipital cortex.

Tom Kirkwood: Could I just ask why the effect should disappear by age 7? If you have such a marked effect early on and it has it has gone by age 7, it is curious.

Christian Drevon: I think there are a lot of things that influence cognitive function, and the effect can be diluted by other factors, including subsequent dietary changes.

Tom Kirkwood: If it is a genuine big difference early in life and then evens out later, there may be some trade-off because the later effects may be disadvantageous. Have you follow-up data that would show whether there is any long-term disadvantage?

Steve Zeisel: Just to help think about why things might be apparent at 4 years and disappear at 7 years: First of all, the hippocampal function tests that you are doing are not very specific; they are general developmental tests which are not terribly sensitive. Secondly, if you have a brain which is maturing – which it is between 4 and 18 years – simple things may be hard to do when you are 4 years old and then, by the time you are 7, you have enough complexity of function that the test is not very challenging and therefore you cannot see the difference. So, you need a more challenging test with a time delay, distraction, or something else. So, normally one is not upset detecting something at 4 years with this type of relatively crude test, only to find it has disappeared by aged 7; there may be hippocampal functions which overcome any simple challenge.

Christian Drevon: I agree; we need more sophisticated tests to assess this issue properly.

Thomas Cederholme: There is not only DHA in the cod liver oil but many other fatty acids of importance for the brain, so what is the optimal composition of a fatty acid supplement, and if it so good, why do we not continue the supplementation during childhood? Would that actually give even better outcomes?

Christian Drevon: In Norway, we do continue to advise everybody to continue with the supplement, and we have a couple of hundred years' experience of this tradition. Cod liver oil contains considerably more DHA than other fish oils, and the main producers of cod liver oil have now removed all the contaminants, but we do not have coherent long-term follow-up studies to assess the comparative advantage of this national practice.

Philip James: Fish-eating in Norway is quite high, is it not? When do children start eating fish?

Christian Drevon: Yes, fish-eating is much higher than in the other Scandinavian countries, but it is only about half the level of that in Japan. Children start eating fish relatively early.

Thomas Cederholme: Is there also saturated fatty acid in the cod liver oil?

Christian Drevon: Yes, but that is not critical at all because there are many other major sources of saturated fatty acids. In arachidonic acid terms, the full-term baby is able to convert linoleic acid to arachidonic acid, whereas if you are a premature baby, you cannot do that, and then arachidonic acid is an essential fatty acid just like DHA.

Ricardo Uauy: DHA in human milk is in the phospholipid fraction, i.e., the envelope for the triglycerides. The DHA content of human milk is affected by the maternal diet. Even vegetarian mothers still have some DHA, this relating to the conversion of linolenic acid to DHA, and that is regulated by the desaturase enzymes. Now we know that women have a limited capacity to make DHA, with the desaturation pathway transforming maybe 1–5% of the linolenic acid into DHA. DHA is highly selectively conserved in the retina and in the cerebral cortex and to some extent in all tissues, but its content is affected by diet. In pregnancy, the DHA increases markedly in the fetal brain towards the end of pregnancy, and this is derived from the maternal conversion of alpha linolenic [acid] to DHA. There are specific placental transporters discovered by Dutta-Roy at the Rowett [Institute]– specifically the selective transporter42– and the fetal liver as early as 26 weeks is actually transforming alpha linolenic [acid] to DHA.43 However, the diet is important, and you have differences in tissue composition dependent on the dietary supply

Philip James: I remember Asim Dutta-Roy's work as showing that with the marked turnover of lipid and fatty acid recycling through the liver, the liver is able to extract the essential fatty acids and, in pregnancy, selectively transport them to the placenta, where a specific transport mechanism ensures their transfer to the fetus. So, it is almost as though there is a special evolutionary pathway to scavenge the DHA from maternal fat and transfer it to the fetus.

Richard Wurtman: The way DHA gets into the brain was worked out by Harvey Lodish.44 An acylase enzyme in the cellular membranes of the blood-brain barrier acylates the DHA. Immediately, DHA is taken up, so this keeps the DHA from diffusing back out and allows its incorporation into diacylglycerol and then triglycerides or phosphatides. DHA has been shown to be a direct receptor agonist on syntaxin45 and induces dendritic spine formation. It is more potent when you give it with uridine, but it does it by itself. So, the range of actions of DHA expands every year.

Philip James: What is the turnover of DHA in the brain?

Ricardo Uauy: In the case of the retina, 99% is recycled, so once you load up your eye with DHA in early life, you do not lose it from the eye. That is very relevant, because we depend on this recycling for night vision. In other tissues, you have some degree of turnover, but DHA is highly conserved in the central nervous system, so the early diet really does affect the brain's DHA content. Changing the brain's DHA content in later life has been observed in mice but not yet, as far as I know, in humans. One last point apropos our meeting here in Sweden is that the discovery that DHA is a gene expression regulator was made by Zetterström's lab,46 with DHA playing a role in activating the retinoid X receptor (RXR), which is a nuclear receptor that functions as a ligand-activated transcription factor.

John Glomset: When humans developed in Africa a long time ago, they survived on green plants, which contained linolenic acid, and were not eating salmon, so the whole system is built on the conversion of alpha linolenic acid to DHA.

Philip James: So, if we evolved with a capacity to make enough DHA from linolenic acid, then why am I being told that the desaturation and chain elongation process is so slow as to be a problem?

Richard Wurtman: In part it is, because we now take in so much n-6 fatty acid that this competes for the same enzyme systems involved in DHA synthesis. That is a consequence of our contemporary diet. In terms of turnover of DHA in the brain, most of it is incorporated into diacylglycerol and then incorporated into phosphatides. So, newly synthesized phosphatide molecules are very enriched in DHA and EPA, but they are rapidly acted on by phospholipase A2, which converts them to the lyso-compounds. Then, whatever fatty acid gets attached back to the 2 position is just a function of what is in the pool at that time. If you measure the DHA content of brain phosphatides, it is fairly constant but a lot less than the DHA content of newly synthesized phosphatides.

Steve Zeisel: I would like to come back to our thinking about genetic variation. We have found a single nucleotide polymorphism (SNP) in the gene for phosphatidylethanolamine N-methyltransferase (PEMT) that makes phosphatidyl choline from phosphatidyl ethanolamine.47 It turns out that when you take the Pemt−/− knockout mouse, it has no or almost no phosphatidyl choline DHA in any of its membranes. That is because phosphatidyl ethanolamine is high in DHA, and when it is methylated to form phosphatidyl choline, it flips to the outer membrane. Humans who have an SNP in this gene have very low phosphatidyl choline DHA in their VLDL and plasma. The interesting thing is that 72% of the Chapel Hill population has one allele for this variant, and 17% are homozygous. So, it could very well be that there are people who have trouble, even when they have DHA, in putting it into the right membranes. So, we should start to look for DHA responders and nonresponders in relation to this allele.

Philip James: So if you are talking about different propensities on the basis of one's genetic inheritance, how do the populations of the world compare?

Steve Zeisel: If I look in a Jamaican population for this specific SNP in PEMT, i.e., a population of African origin, only 22% of the population have the one PEMT allele, and 5% are homozygous. Chinese Asians in the Cornell database seem to have a prevalence of the PEMT allele, which is halfway between the Caucasian and the African prevalence. But I would say there are fairly common SNPs in the desaturases that accompany essential fatty acid elongation, so there are people who are slower or faster in generating DHA, and I think again that information is so easy to collect now with the SNP chips that if you were doing human studies, you should be collecting the genetic data. So, one day, when we identify responders and nonresponders to DHA, you can ask the question, “Were they different genetically and metabolically?”

Christian Drevon: On the topic of alpha linolenic acid and the optimum supply of DHA, I do not think anybody has shown in children or adults that you really need DHA or EPA by itself. If you get alpha linolenic acid, as John Glomset is saying, you are fine. However, I think in our environment it is best to have, in addition, the marine very long-chain omega-3 fatty acids, e.g., 1 to 2 g a day. Then, when dealing with fetal development in special groups, e.g., premature babies, it is a totally different story because they have a very low capacity for essential fatty acid elongation and desaturation.

Richard Wurtman: Given our intake of n-6 and the competition for uptake into brain between the n-6 and the n-3 fatty acids by this mechanism, together with the competition for acylation and competition for incorporation into DAG [diacylglycerol], it does look as though a supply of marine-derived fatty acids is valuable.

Ricardo Uauy: The need for DHA is evident during early growth in both term and preterm babies. After that, in vegetarian populations, there is no functional defect in these linolenic-acid-consuming populations that do not take up DHA later on in life. Whether their aging is different, we do not really know, but from the standpoint of their needs, the alpha linolenic acid provides the DHA needed in early life.

Greg Cole: We have clinical trial evidence that fish oil supplements reduce the risk of sudden cardiac arrest, at least in the Western populations, so in terms of our risk of cardiovascular disease, I need more DHA or omega-3 fatty acid than I am getting from alpha linolenic acid. In the context of the development of Alzheimer's disease, where one finds amyloid precursor protein (APP) and APP-derived amyloid-beta (Aβ) peptides causing synaptic dysfunction, the levels of activated GIVA-PLA2[group IV isoform of phospholipase A2] in the hippocampus are increased, and with the activation of phospholipase A2 GIVAthere is more release of arachidonic acid. In experimental mice, inhibition of GIVA-PLA2 diminishes Aβ-induced neurotoxicity and protects the mice from deficits in learning and memory, from behavioral alterations, and from premature mortality.48 This means that inhibiting GIVA-PLA2 may be beneficial in the treatment and prevention of Alzheimer's disease. So, the competition between n-6 and n-3 can become relevant to the pathophysiology of Alzheimer's disease.

Philip James: So, when one now considers intervention studies with DHA, are you confident that in terms of mental function, DHA has been seen repeatedly to be beneficial?

Greg Cole: There is epidemiological evidence, small clinical trials, and animal model data. Now we need large clinical trials to show DHA's efficacy. In terms of animal model data, you can see remarkable protection, but we have to wait to have clinical trial data to show what it really does in humans.

Thomas Cederholme: Probably the largest human studies so far involved our gathering about 200 patients with pre-existing mild-to-moderate Alzheimer's disease with Mini-Mental State Exam (MMSE) scores above 15 and randomizing them to have either 3 g of mainly DHA for 6 months or placebo.49 After 6 months, we opened the study and then everybody received the DHA supplementation. When we analyzed the data, we were disappointed, as there were no effects on cognition in the whole group, but when we did subgroup analysis, we saw that in the group with the mildest forms of Alzheimer's disease, they actually had a slower decline in MMSE. That was probably one of the first longer term studies showing some effects, but we still think that we started late in the disease's progression. So, we probably have to start supplementation much earlier.

Vidar Gundersen: The inhabitants of Greenland are fed on a diet very rich in fish oils, so I wonder whether we know anything about aging and age-related diseases in that population.

Christian Drevon: I have been there and seen their lifestyle, and it is amazing – especially the amount of alcohol consumed. At the weekend, their drinking is even heavier than that seen in Norway! Although they do not get coronary heart disease, they do not live long because of the effects of alcohol with trauma, violence, and acute intoxication. There are also only a few thousand Greenlanders or Inuits now living on a traditional diet, because most of the population gets its food imported from Denmark, so I think it is impossible to draw conclusions based on this population.

Helga Refsum: In Western Norway, where everybody more or less eats a lot of fish, we investigated the relationship between cognitive function and fish intake.50 There were actually only 40 out of 4,000 who did not eat any fish, and they were so special that we could not compare them with the rest of the population. But we could look at dose-response relationships, and there was a very sharp dose-response relationship between cognitive function and fish intake. These people were 70–73 years of age, and we did find the effect in all six different cognitive function tests. We also found that half of the population used a good dose of cod liver oil; they had lean fish and fatty fish, and the dose-response was as good for lean fish as it was for fatty fish, but the relationship was not as strong for cod liver oil. So, it looks as though the intake of fish, rather than cod liver oil, is important. Then, when you go back to the other epidemiological studies, you find the same thing: it is the overall fish intake that is good for you rather than fatty fish intake as such. So, it is perhaps something to do with eating fish. It could be the unsaturated fatty acids, the vitamin D, vitamin B12, taurine, or something else, or even the effect of the fish replacing another food item which limits the mental deterioration.

Ricardo Uauy: We have a paper with Dr Martin Prince involving a cohort of 15,000 people over 65 years of age from 10 countries, including Cuba, Mexico, and China.51 Out of the 10 countries, in all the countries except India, there was a significant effect of eating fish. Eating fish at least once a week reduced the risk of dementia, based on a standardized questionnaire, by 16%. In India, there was no effect.

Philip James: Because In India they do not eat any fish?

Ricardo Uauy: There were some Indians who reported eating fish, but that sample size was smaller than in other countries. We are in the process of completing a controlled trial. We studied 800 people in the United Kingdom, 70 years and over, and gave them 500 mg of DHA plus 200 mg of EPA, which is the equivalent of consuming oily fish twice a week. Up to now, we have only the baseline data; the protocol has been published52 and the study will be completed very soon. We find that there are significant effects in terms of memory: the Californian Verbal Learning Test, global memory, processing speed, and executive function are proportional to the amount of fish consumed at baseline. Adjusting for age, sex, and the age at which they finished their education, the values still remain significant, but the coefficients – the strength of the relationship – are much weaker, although they remain strong. Now in the prospective phase, we have not analyzed the separate groups, but we have analyzed the fatty acids in the blood, and we find that we are getting a major effect of DHA, at least in the chemical measurements that we have made. The only problem, though, was that we had to reject a third who were already consuming fish oil. And the rate of functional decline that we were expecting to find in the California Verbal Memory Test was, in fact, slower than predicted because this is a healthier group compared with the general population.

Philip James: You talked about sex differences?

Ricardo Uauy: Women are much more intelligent – they have better scores!

Steve Zeisel: I just want to emphasize an obvious point: the quality of fish depends on what they eat, and with fish farming becoming prevalent, we need to recognize that they are fed corn, with possibly no health benefit, whereas fish feeding on algae in the Atlantic may be very different in their composition.

Ricardo Uauy: Fish actually require omega-3, so you cannot give them corn, but you can give them soy. In fact, they are now often fed soy and then, 2 weeks before they are put on the market, they are given some DHA derived from fish meal from Peru. Normally, of course, algae are important sources of n-3 fatty acids for fish.

Cutberto Garza: Helga Refsum makes the point that studies often relate mental outcomes to previous documentation of dietary types. Then, Dick Wurtman has made the point that if you combine two or more of these nutrient components, you get better effects. So, if you go back to look at the data, either epidemiologically or perhaps in a more controlled way, then the strongest relationships are always with diets rather than single nutrients. When we try to reduce the impact of a diet to a single component, we may also come up with a totally unexpected outcome, the beta carotene trials − based on earlier dietary relationships − being one of the best examples. So, it is important that we understand, at least from a nutritional perspective, that specific nutrients and food are often distinctive in terms of the outcomes.

CHOLINE AND B VITAMINS

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Steve Zeisel: We need to think about early events in life when we are considering neurogenesis. I will illustrate this point with some animal nutritional experiments involving choline as a methyl donor for the one-carbon pool turnover of the body. When we feed mice from day 12 to 17 of pregnancy – the time of maximum hippocampal neurogenesis in the mouse – and assess the results on day 17, then we can consider the effects of a high-choline diet (with about 4 times more choline than normal), a controlled diet, and a choline-deficient diet, which has about a tenth of the normal intake. We find that in different regions of the hippocampus, neurogenesis depends on the choline intake. We also find a halving of apoptosis in the supplemented group, whereas in the deficient group, apoptosis rates have doubled.53 Colleagues at Duke University have also published recently a series of studies in mice after birth and up to 2 or more years of age.54 The mice were born and then cross-fostered to a control mother and fed a normal choline diet for the rest of their lives. Only for 5 days during gestation were their mothers' diets manipulated. They were then assessed for their performance in a maze experiment. The normal control animals got worse and worse with age, but the offspring of the choline-supplemented mother did not lose any function with aging; i.e., they had the same maze performance throughout: the offspring with a supplemented fetal exposure had a 30% better maze performance at 2.5 months of age, and this was maintained throughout life. We have presented data suggesting this is an epigenetic effect, with CpG base islands in the promoter regions of a number of genes controlling stem cell proliferation being methylated. This effect occurs during a very narrow window in the time of precursor cell proliferation before differentiation. So, during critical periods in development, dietary methyl-group intake (of choline, methionine, and folate) can alter DNA and histone methylation, which results in lifelong changes in gene expression55 Once differentiated, the cells cannot be changed in this specific epigenetic way. Now, when considering potential neurogenesis later in life, we need to recognize that about 95% of neurogenesis in humans probably occurs before 4 years of age in the hippocampus, with probably 1–2% of the neurogenesis continuing to occur in mid or late life. So, it is possible for these dietary-induced epigenetic effects on neurogenesis still to affect cell numbers later in life, but we have never looked at that. It also remains unclear how these epigenetic effects might mediate the preservation of memory in the animals fed choline during fetal life.

Ole Petter Ottersen: Is this memory test dependent on intact olfactory function? Because obviously, the neurogenesis could relate to the olfactory function rather than to the hippocampal function.

Steve Zeisel: No, the memory testing is done in two ways, and one of the tests involves a visual test: they have to recognize where a hidden underwater platform is.

Richard Wurtman: Is it possible to look at a dose-response effect of early choline supply on later memory?

Steve Zeisel: We have only done limited animal studies. However, in Berkeley, California, the March of Dimes found that women in the lowest quartile for choline intake had four times the risk of having a baby with neural tube defects as women in the highest quartile in California.56 The lowest intake was about 250 mg/day and the highest intake was about 650 mg/day, so there is a twofold variation in Berkeley. Probably, between developing countries and the highest intake, there could easily be a fourfold difference.

Philip James: Where in dietary terms can I get choline and the other components we are considering, e.g., uridine?

Steve Zeisel: The USDA [US Department of Agriculture] has a database on their website. A vegetarian in the United States has wheat germ as their major source of choline, and probably in other places it is germ-like plants which have a lot of membranes; you will not get much choline from plants otherwise. The major source is meat and eggs for nonvegetarians. Human breast milk is also a very rich source.

Richard Wurtman: Most of the experimental studies described, including my own, are really pharmacological studies in which we have administered relatively large, even huge doses. Or, alternatively, we have put animals on zero intakes of a nutrient. I think you need to do these types of studies to establish that there is a relationship between the variable we are studying and the factor measured. However, these approaches do not really tell us very much about what happens normally. For example, if we knew that plasma uridine levels varied over a fivefold range (which we do not know!), then we could perhaps speculate about whether or not the bolus of nutrients is giving an appropriate input in terms of an increased plasma uridine level within the documented normal range. Thus, it might be better to feed somebody something after lunch to increase the levels more effectively, but we have no such information. We do know if plasma choline levels vary over a three- or four-fold range. The MIT [Massachusetts Institute of Technology] students in our study eat as much omelette as possible, but the physiological significance of these intakes is anybody's guess. I think the same thing tends to be true with DHA tests. I think our understanding will be clearer when we undertake dose-response studies within the appropriate physiological range.

The problem with uridine is similar: we do not know about normal variations in plasma uridine levels or the effects of various foods on plasma uridine. We tried giving MIT students a lot of herring or anchovies, thinking that the RNA would be a good source of uridine, and we never really saw an increase in plasma uridine. One gets a big increase when we give uridine monophosphate as such, but the amount given in uridine-equivalent terms is a lot greater than that obtained from foods. Liver makes uridine, so it is likely that uridine is, like choline, provided endogenously and also as a food constituent.

Ricardo Uauy: Human milk is loaded with uridine, but not processed cows' milk, but we have to remember that Hamish Monroe demonstrated quite well the salvage pathway is a way of ensuring recycling of uridine, which is important for rapidly dividing cells and also for enterocyte function.

Irv Rosenberg: An issue which emerged from Dick Wurtman's exciting data is the concept that we need to identify those enzymes involved in nutrient metabolic pathways that have a low affinity and which therefore might be responsive to increasing the amounts of their substrates with an increase in the flow-through of the pathway as substrate concentrations rise. So, that is at least an additional paradigm of thinking about nutrient deficiency or subclinical nutrient deficiency or the effectiveness of nutrient supplementation in general. How much do we know about a profile of enzymes and nutrient metabolic pathways so that we can identify those that have low affinity and might be responsive to increases in their substrate precursors? Is that really another way of looking at the question of nutrient intervention?

Steve Zeisel: One problem with looking at these enzyme pathways is that many pathways have a rate-limiting step at their end. So, for instance, in phosphatidyl choline synthesis by the Kennedy pathway, the cytidyltransferase is highly regulated and phosphorylated to make it inactive in the cytosol and then dephosphorylated to go to the membrane where it functions. As with many biochemical steps, the last step is the critical step that is tightly controlled, and it would make it very hard for me to understand how loading up ahead of that step makes any difference, unless you have got a pathology or abnormal neuron that is below whatever it is trying to regulate. Then you may return this step up to normal, but the cytidyltransferase step is not limiting. So, I think we have to look very carefully at where the squeeze or constraint point is in metabolism. Once identified, the issue is whether one can alter the flow through this restricted step in the pathway. If you cannot push substrate through it, there is no point in loading up with nutrients before this step.

We have recently found that endothelial precursor cells react the same way as our neural precursor cells to changing choline supply. We can upregulate their proliferation epigenetically in almost the same way, so probably both choline and folate supply affect methylation and interact. However, we have given folate to our choline- deficient animals and cannot reverse the apoptosis, but we can moderate some of the effects on proliferation.

Richard Wurtman: Specifically in the case of phosphatidyl choline synthesis, all three steps can be rate-limiting. Consider, for example, the way nerve growth factor enhances phosphatide synthesis at the last step: DAG levels increase about three- or fourfold.

Jacob Selhub: I think what Richard is saying is that we are dealing with a black box. The whole system of one-carbon metabolism, the correlation with choline, phosphatidyl choline – all this stuff; it is all a black box. So, the strategy, if you are to intervene, is not to intervene with one single nutrient, but you may need a slew of everything. Steve said the rate-limiting step may be at the end of the process, but if you have more substrate at that step, you will still get more product.

Ricardo Uauy: One of the problems is that the optimal dose may not just be an issue of adding up nutrient levels, I would like to have explained what the underlying mechanism is when B12 levels are low and folate levels high, leading to a worse outcome.

Jacob Selhub: The antagonism between high folate and low B12, may reflect the impact of folic acid per se rather than the folate as such.

Helga Refsum: I just wanted to say in terms of vitamin B12/folate, we have to remember that methionine is critically important in regulating more or less everything. And S-adenosyl methionine levels will be critical in determining a shift in folate use away from DNA synthesis to homocysteine remethylation. It is also affects whether homocysteine is either remethylated or transulfurated, and we know that methionine is so critically dependent on the diet and its protein content. So, protein restriction involves methionine restriction. Very many experiments have actually been done with methionine restriction, and you will get exactly the same effect with methionine restriction as you get with calorie restriction. So, methionine restriction may be the mechanism for much of the effects of limited food calorie intakes. At least in rat experiments we have shown that if you do methionine restriction, you actually prolong life – with less diabetes – through the same mechanism. So, I think that methionine is exceptionally important but that cysteine is the really critical factor.

Jacob Selhub: Helga is correct about the methionine restriction inducing the same effect as a low-calorie diet. Methionine is an essential amino acid, so this normally means that it is not synthesized in the body. Yet methionine is synthesized in the body, but we need more, and there is a balance how much is needed in relation to other substrates and nutrients.

Patrick Stover: If you look at the folate pathway, the most efficient folate metabolism is found in a tumor cell, which has about half the folate of the normal cell. Steve Benkovic has shown that when you have active proliferation and need to make nucleotides and require S-adenosyl methionine, the enzymes involved come together in a complex and channel the folate so that it does not diffuse away.57 So, sometimes you get more efficiency at subsaturation when you have complexes and channeling. The other point is that when you put animals on high levels of folate, you get downregulation of the receptors and downregulation of some of the enzymes because of the higher folate concentration. When you then switch them to a deficient diet, it takes a long time to adjust both transporter expression and some of the enzyme expression to match the new levels of folate that are available, so these systems are very responsive in terms of availability of the nutrient. So, there can be trouble when you have variations in diet because the system does adjust to the availability of the cofactors.

Aron Troen: There is a very interesting study using stable isotope tracers looking at differences in flux during choline deficiency and folate deficiency, and although they both impact methylation end products, the metabolic fluxes are very different.58 So, although we may have a common measurable index such as plasma homocysteine, this does not necessarily specify the nature of the underlying metabolic fluxes.

Susan Folstein: Does it make any difference how old the mice are when you do your experiments?

Aron Troen: We are planning studies on an aging mouse cohort; presumably, in the older mice, where neurogenesis is impaired, the issue is whether or not folate amplifies the animal's vulnerability.

David Smith: Folic acid provides a very good example of a rate-limiting step because folic acid only enters the folate cycle by being reduced through dihydrofolate reductase, which has an extraordinarily low activity in humans. So, you might think you are flooding the body with folate when you give folic acid, but you are not. Not only is dihydrofolate reductase of low activity in humans but it is also rather variable from individual to individual, and there are polymorphisms which of course Jacob has studied. So, the entry of folic acid into the system is going to be individually variable, and therefore we have to be very careful predicting the effect of loading doses of nutrients.

Joshua Miller: We learned from Dick Wurtman that we not only have to include uridine, DHA, and choline but also the B vitamins: B12, folate, and B6. So, the balance between the different pathways involved in metabolism not only requires the nutritional precursors but also the cofactors involved in those reactions. Another general point is that when we are giving nutrients as precursors, we are often affecting the biochemistry, but if we already have lost some brain structures, then it is difficult to remedy the problem. So, we need to consider the importance of preventing dementia/Alzheimer's disease rather than treating it. We have been running the Sacramento Area Latino Study on Aging, also known as SALSA,59 for many years now, where the baseline collection of samples and assessments for this study occurred just after folic acid fortification was instituted in the United States in 1998 – actually, it probably started in earnest in 1997. In this folic-acid-fortified population, an analysis of data from the National Health and Nutrition Examination Survey (NHANES) had indicated that in older adults exposed to folic acid fortification, the combination of low serum vitamin B12 and elevated folate was associated with higher concentrations of homocysteine and methylmalonic acid and higher odds ratios for cognitive impairment and anemia than the combination of low vitamin B12 and nonelevated folate. In our study we found no differences in modified MMSE, delayed recall, and depressive symptom scores between the low vitamin B12 and elevated-folate group compared with other groups. However, plasma homocysteine levels were associated with domains of poorer cognitive function. Then when we looked at folate itself in blood, it was also associated with different domains of cognitive function, which were independent of homocysteine levels.60 A lot of people think of folate supply as simply lowering homocysteine levels, but in our follow-up study about 4–5 years later, we found that the incidence of cognitive impairment (which was defined as either outright dementia or what is called “cognitively impaired but not demented”) was now not strongly associated with the baseline plasma folate levels. So, it seemed as though the folic acid fortification has had an effect on folate-related dementia or cognitive impairment. Those people who had low B12 and high homocysteine were the ones who seemed to be getting the dementia.61 We are taught that folic acid can mask B12 deficiency, and these folic acid studies are revealing that B12 is the next vitamin that needs to be studied intensively.

There are also data about vitamin B12 deficiency amplifying the problem of inflammation. This evidence mainly derives from Scalabrino's lab in Milan, where they find that B12 deficiency is associated with higher TNF-α[tumor necrosis factor alpha] levels and excess inflammation.62 Given that inflammatory disease is involved in Alzheimer's disease, B12 deficiency may be exacerbating the inflammation.

I also want to make the case for vitamin B6. This is a very understudied vitamin, particularly as it relates to the brain, but it is very important for neurotransmitter synthesis. We have found in our SALSA population that there is a very high percentage of people who have a low [level of] pyridoxal phosphate, i.e., the active form of B6, circulating in their blood. We also looked at Alzheimer's patients versus controls, and we found that about 20–25% of our Alzheimer's patients had low pyridoxal phosphate compared with controls, where it was only about 9%. The other thing we have shown is that in a subset of the SALSA population, we did fluorescent deoxyglucose PET [positron emission tomography] studies to look at glucose metabolism in different locations of the brain. We found that in those people who had cognitive impairment, there was a direct association between B6 and glucose metabolism in specific sublocations of the brain.

I want to add a point about “nutritional pharmacology,” i.e., how nutrition affects the usefulness of drugs. One example is B6 and the drug memantine. Memantine is used for Alzheimer's disease, and if I remember correctly, it affects the release of glutamate, and B6 is necessary for the conversion of glutamate to GABA. If you are low in B6, then perhaps you cannot metabolize glutamate very well. So, perhaps the efficacy of memantine could be affected by your B6 status.

Aron Troen: Briefly responding to the nutritional-pharmacological interaction, there is a nice study by Manganelli et al.63 on movement disorders looking at the response to cholinesterase inhibitors and cognition in Parkinson's patients showing that homocysteine is a predictor of response. Those with high homocysteine levels respond very well to cholinesterase inhibitors; if you have normal homocysteine, the response is much more muted. One can look at the interaction between cholinergic pathways, choline availability, and so on. So, there is good evidence for that, clinically. With regard to B vitamin requirements in the aging brain, one of the things that needs to be considered – and this is true of all nutrients – is that aging changes nutritional requirements, and the brain is particularly susceptible because of the impairment in the specific transport of nutrients into the brain. So, back to the vascular factors: the choroid plexus, which is the locus of transport of many water-soluble vitamins, including folate, into the brain, ages remarkably as a function of vascular aging. In fact, 60% of elderly individuals have calcification of the choroid plexus – it is a very common finding on CT, and when I was working with David Smith in Oxford, we dissected 150 choroid plexi, and one can readily feel the choroid plexus stones, which are granular and calcified. Nobody knows how that affects transport, but we do know that CSF production decreases approximately by half over the course of our lifespan, so even if peripheral B vitamin levels are ostensibly normal, we know very little about the actual brain concentrations.

FOLATE AND NEURAL FUNCTION

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Patrick Stover: I'd like to make two points relating to the role of folate in, first, maintaining brain integrity and, secondly, in the possible genetic selection for an aging phenotype.64 One of the problems with studying the relationship between the mechanisms responsible for folate deficiency is that the deficiency impacts two important pathways: the methylation pathway, i.e., the synthesis of S-adenosyl methionine, which relates, for example, to chromatin stability and gene expression, and secondly, all the consequences of an impaired nucleotide synthesis, with inappropriate incorporation of uracil in DNA and therefore the possibility of elevated mutation rates. The synthesis of nucleotides is required for both DNA synthesis and repair, and both are therefore affected by folate deficiency. However, one needs also to understand the different distribution within cells of the different components of the one-carbon pool pathways, their dependence on folate,65 and the way in which different allelic properties of the important enzyme systems affect the fluxes through specific pathways. Thus, the hydroxymethyl group of serine enters the folate-activated one-carbon pool through its tetrahydrofolate (THF)-dependent conversion to glycine and 5,10-methylene-THF in a reaction catalyzed by the enzyme serine hydroxymethyltransferase (SHMT). SHMT is present in both the cytoplasm (cSHMT) and mitochondria (mSHMT), and the SHMT isozymes are encoded on distinct genes, SHMT1 and SHMT2, respectively. We had already shown that if we look at SHMT-deficient mice, lacking either one or two alleles, encephalopathy and spina bifida results. This is the first mouse model where a folate metabolic gene has been knocked out and it has resulted in a neural tube defect, indicating that it is probably the thymidylate pathway that is important in maintaining genome synthesis rather than the homocysteine remethylation cycle, which most people consider important. So what about aging? With these same SHMT knockout mice, we have assessed neurogenesis in aging with the three different genotypes of SHMT deficiency, the homozygous, the heterozygous, and the wild-type, all fed either 2, 4, or 8 mg per kilogram folic acid. At both 3 months and 14 months, we assessed hippocampal function using a classical fear-conditioning test and then assessed whether they have retained the warning signal linkage with the undesirable shock 24 hours later. The mice with the lowest capacity to make thymidylate are not able to make the association between that sound and the shock by freezing their movements. So, this shows the importance of genome maintenance in an aging phenotype related to memory.

There are plenty of polymorphisms in genome maintenance which are risk factors for various conditions in man, e.g., spontaneous abortion and miscarriage, and those same genotypes that optimize fetal development will probably be needed to optimize the function of stem cells in neurogenesis later in life. So, it is very possible that the selective pressures in utero also contribute to a healthy aging or longevity phenotype; that is one of the hypotheses that we are testing.

David Smith: So, your conclusion was that in this knockout model, thymidylate was the determining factor for NTDs [neural tube defects]. How does that fit with the human NTD MTHFR TT genotype, where you would predict a very different mechanism?

Patrick Stover: Jacob Selhub has shown that this MTHFR polymorphism lowers cellular folate concentrations up to 40%, so the polymorphism creates a functional folate deficiency. So, you can have a genetic susceptibility to the metabolic defect with folate deficiency.

Irv Rosenberg: Patrick, you do get uracil misincorporation in a number of models of folate deficiency as well as even in some other deficiencies. Do you also get it in the presence of low B12 and high folate?

Patrick Stover: We looked at all potential readouts of the one-carbon cycle: methionine, adenosyl methionine, and microarrays of protein expression and CpG methylation. We saw nothing with this model except elevated uracil levels in the DNA.

Philip James: This issue of appropriate folate levels is important because Ranjan Yajnik from Pune in India finds that India has a prevalence of 75% of B12 deficiency in the population. In beautiful cohort studies involving the monitoring of women before and throughout pregnancy and evaluating the offspring into their teenage years, he finds that the pregnant women who have the highest folate and the lowest B12 have the most abnormal small babies. They are small but fat, with very little lean tissue, and even at birth show signs of insulin resistance, which is amplified at 4 and at 8 years of age.66

Jacob Selhub: We have also seen that in the US NHANES study, some people who consume 5 mg a day of folic acid also have low B12, and some are glucose intolerant, too.

Helga Refsum: In Yajnik's group in India, a decline in short-term memory and sustained attention has been found in 9-year-old children born to mothers who were vitamin B12 deficient at 28 weeks of pregnancy.67 Did you measure methionine in these mice?

Philip James: Patrick, when we are talking about folate and so on, you are talking about replicating cells as well as any potential effects on methylation. What about cells like the immune system? Do we affect the immune system when we have deficiencies of folate, B12, or choline?

Vilhelm Bohr: Uracil accumulation in certain DNA regions is associated with hypermutation in the lymphoid B-cells, so that is a very important process. There are also studies looking at uracil glycosylase activity with age, but the results depend on which tissue is assessed; generally, it declines with age.

Kaare Norum: Have you seen anything similar to the Indian findings in the United States, where there is folate supplementation and vegetarianism with low vitamin B12 intakes?

Irv Rosenberg: We obviously do not have anything close to the prevalence of low B12 among the US population or the pregnant population in the United States compared with India; body compositional issues have not been assessed as in the fascinating data from Pune.

Helga Refsum: We have debated it extensively in the United Kingdom. We have a large Indian population in the United Kingdom with a low B12 intake, and we have raised this problem frequently, asking that investigations be done in terms of B12 status and folic acid, but these studies have not yet been done.

Aron Troen: It is important to distinguish between classic clinical B12 deficiency and a low B12 status in the natural range of the population. David's work on the risk for brain atrophy occurs in the lowest two tertiles of B12 deficiency, meaning about 60% of the population are at risk. So, the importance of B12 deficiency does not only apply to cases of frank deficiency.

Ingmar Skoog: I am not sure how relevant some of these factors are in relation to brain aging. What we know very clearly is that a lot of different factors throughout life – vascular factors, B12, and other factors affecting cellular wear and tear – influence brain aging.

David Smith: Why have we not discussed the famous Rowan Kahn idea68 of usual and successful aging? If you look at brain function, we see a decline in function over time. So, if you follow a cohort of people over 20–30 years, the average cognitive scores will decline. But if you look in detail at this cohort, you will find that some people do not show any evidence of a decline. In fact, in Optima we have some people we have been studying for 20 years, and their cognitive scores are still the same as they were when they came into the study. They are amazing; they are in their nineties now, having been recruited in their seventies. So, what Rowan Kahn suggested was that there is something called successful aging, which few of us, unfortunately, achieve. Successful aging depends on a mixture of intrinsic and extrinsic factors which modify their performance in the brain or in other tissues. Vascular factors, vitamin status, and atrophic processes are all involved.

Kaare Norum: Just a small historical note: about 50 years ago, as a young doctor in a psychiatric hospital, I noticed that lots of the elderly patients coming in for acute psychiatric disorders were vitamin B12 deficient, and when we treated these patients with vitamin B12, they improved remarkably; this was my first clinical paper!69

Marshal Folstein: Although there is no question that severe deficiencies of B12, thiamin, and niacin can be partially reversed and you can see clinical improvement, there are data suggesting that the subjects are also left with residual deficits, both in their myelin and in their diencephalon in the hippocampus.

FOOD FORTIFICATION

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Jacob Selhub: I think one of the lessons that we have from the food fortification in the United States is the fact that it was a single fortificant, whereas from an understanding of standard biochemistry, we should be considering the whole pathway and a range of nutrients. When Dick looked at the pathway of choline synthesis, he added uridine and considered the whole pathway. If you add the single nutrient, you get imbalance, and that's where the problem exists. In the case of folate, when they give folic acid without B12, all of a sudden there is an imbalance.

Irv Rosenberg: We do not have monofortification in the United States. We had a number of B vitamins and other things in the fortification of flour before we ever added the folic acid. You may be right about our inability to predict exactly the effect of folic acid, given its interactions with B12 and the issue of B12 deficiency, but I do not think that fortification was confined to folic acid. To try to estimate the effects of multinutrient fortification is complicated. The other challenge relates to David Smith's suggestion that most of the enzymes in the folate-dependent pathways are low-affinity enzymes. I am not sure what the evidence for that is, compared to other pathways, but even if it were true, it still is not entirely clear to me how to predict whether adding more substrate to a low-infinity enzyme in a pathway is going to result in a particular product. Dick Wurtman's suggestion that we might understand some of the effects of nutrients by understanding their effects on pathways is a very intriguing and interesting one. We need more information about the regulation of these pathways and how much you can actually accomplish by adding substrate to one or other sites within the pathway.

CALORIC RESTRICTION

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Irv Rosenberg: In my discussions with Tom Kirkwood relating to the need to think about aging not in terms of the whole organism but in terms of different systems and physiological effects, he rather forcefully rejects the idea that the effects of calorie restriction seen in animal studies applied to increasing longevity in Homo sapiens. If one induces calorie restrictions, do you get different kinds of metabolic responses in different tissues, and what changes occur in the brain?

Christian Drevon: I think there is a lot of evidence that if humans are restricted in their energy intake, a decrease in blood pressure, cholesterol levels, triglycerides, free fatty acid levels, and the number of inflammatory cells in different organs occurs, i.e., exactly what you see in animal studies as well. So, to be so firm about caloric restriction does not seem to be supported by the data.

Vilhelm Bohr: Calorie-restricted mice do not get any age-associated disease and no or many fewer cancers, so it is a major intervention to consider, and it connects up with all sorts of metabolic pathways which have been studied across the animal kingdom, from nematodes to mammals. Caloric restriction may or may not work in humans; it is true that the primate studies so far do not indicate a big difference in lifespan.

Mats Rudling: I think that caloric restriction in humans could be quite important because studies by Walford et al.70 in the Biosphere 2 project showed that people sealed in the Biosphere and restricted in calories for 2 years originally had normal blood pressure, blood sugar, plasma cholesterol, and hormonal levels, e.g., insulin and thyroid hormones, which all fell, as well as the marked decline in weight. These classic risk factors are well recognized to have quite a large impact on survival. So, I think that caloric restriction in humans should really have similar beneficial effects as in rodents. Although I do not believe that we can double or triple our lifespan, we can certainly improve it.

Thomas Cederholme: Caloric restriction is not the same as starvation. In the animal studies, the studies prevent ad libitum feeding and keep weight stability from youth up to old age. So, caloric restriction is a concept that is in line with the general nutritional recommendations.

Vilhelm Bohr: I consider caloric restriction to mean 30–50% less food. However, this is not compatible with a reasonable life for a human being, and that is why Walford and the others gave up in their Biosphere experiment: because it was simply no fun. We are not talking about a simple restriction but cutting down the food to a bare minimum, so it is not really compatible with a normal lifestyle.

Philip James: There is an enormous literature showing that we have brain networks that are duplicated to ensure that we do not cut down our food intake readily. Even after overfeeding and becoming obese, it is extremely difficult to get people then to drop their weight by more than 10% because the body immediately switches on “semi-starvation processes,” even though the individuals are still obese.

Kay-Tee Khaw: The caloric argument is a bit simplistic in the absence of external things such as extreme physical activity and infection, because a lot of the animal experiments are on animals in very restricted environments and not in the free-living situation where they are exposed to lots of stimuli. We must not be too simplistic in our argument; when Ranulph Fiennes went to the Arctic, he was eating 6,000 kcal, 60% fat, but his blood cholesterol still dropped, as did his weight. We also know that, about 10 years ago, trials were done where people were given identical diets spread out in two meals or seven meals; those given the food more frequently had lower lipid levels and lost weight compared to people on identical caloric diets who had only two meals a day.71 This has also been replicated in the free-living observation data in several populations, including ours.72 So, even though people may be eating more, if they eat more frequently, their metabolic profile is lower, and that is related to the insulin secretion patterns as well. So, I think we have to be much more careful when we talk about simplistic things like caloric restriction. Are we hunters and gorgers, or grazers?

Jon Storm-Mathisen: This frequent eating has been practiced in rural communities in Norway, where they would eat not very much but very many times a day. In relation to Catholic restrictions on eating particular foods or the Muslim tradition of fasting, there are some animal studies where food is removed for one day a week, and this had the same effect as restricting the calories in percentage terms.

Philip James: Kay-Tee's point is that the flux relating to frequency of inputs can actually have quite an important part to play in terms of insulin responses and so on, and that in turn alters metabolism.

Aron Troen: There is some very important work that came out of Linda Partridge's lab in the drosophila model, looking at the composition of the fruit fly diet which promotes longevity. Using isocaloric restriction but varying the protein and sugar composition of the diet, it was found that the nature of the protein component promoted longevity, and sugar restriction did nothing.73 That was attributed – at least speculatively – to the effects on nutrient-sensing and an interaction with IGF.

Cutberto Garza: Kay-Tee's point is important because those effects are better understood than we all realize. Experiments many years ago as well as newer studies highlight that the caloric cost of exercise of equivalent strength is different in the fasting or postprandial state.74 In the postprandial state, the same work required much more energy than in the preprandial state. So, we used to refer to it as “specific dynamic action” that was influenced by exercise; now it is realized that it is a sympathetic effect which is amplified in those who habitually exercise.75 So, if you are generating heat and you are in the North, I think one can understand those food practices being conducive to heat generation, and therefore you stay warmer if you eat frequently and remain active versus the preprandial or inactive state.

Bo Angelin:[James] Levine from the Mayo [Clinic] emphasizes the importance of nonexercise activity thermogenesis.76 Some of us are sitting still and others are continuously moving, and these differences can make a lot of difference in calorie expenditure over time. Mats and I have been very interested in fibroblast growth factor 21 (FGF21), with PPARα controlling this hepatic factor which is secreted and has remarkable metabolic effects. Secretion is initiated very strongly by short-term fasting in a mouse, so if you keep it without food for 8 hours, you get a tremendous response, whereas in humans, you have to deprive them of calories for 7 days before you see a small shift upwards.77 That is why I think caloric restriction, meaning being fed every other day, is for a mouse something completely different when compared to us being fed every other day. The differences probably relate to species differences in metabolic rate, and we have to be very careful about extrapolating data from mice to men and vice versa.

THE IMMUNE SYSTEM

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Ole Petter Ottersen: We have not addressed a major system in our body which is normally responsible for repair, i.e., the immune system. There has recently been a sea change in views on the relationship between the immune system and the brain. Much of this research is controversial, but [Michal] Schwartz78 has beautiful work on the relationship between T cells and neurons. She recognized that it was commonly believed that the nervous system functions optimally without any immune intervention and that any immune cell infiltration to the CNS is a sign of pathology. However, since the immune system constitutes the body's major defense and repair mechanism, it seemed unreasonable that the CNS would have completely lost the need for assistance from this system. So, she claims that the first system to age is the immune system, and this means that the CNS, the brain, will also be at risk of aging. She claims that there is a sort of protective autoimmunity in which T cells are involved in an interplay with the neurons. T cells mediate their effect, at least under pathological conditions, by controlling the recruitment of blood-borne monocytes, which play a crucial local role that cannot be replaced by the resident microglia in the brain. Boosting a T-cell response specific for brain proteins with a careful choice of antigen, its carrier, and care over timing and dose, she thinks, should be considered as a way of augmenting a physiological repair mechanism and restoring the cellular equilibrium in the brain needed for protection, repair, and renewal. Michal Schwartz and several other groups' claim is that there are specific antigens on the neurons that are targeted by the T cells and that this interaction is protective.

PHYSICAL ACTIVITY

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Jon Storm-Mathisen: There is clear evidence that exercise has a beneficial effect on the brain, but it is not so clear exactly how exercise influences the brain and involves neurotrophic factors. Could exercise be improving the vasculature?

Jon Storm-Mathisen: Exercise, which is so good for everything, is also supposed to influence telomere length.79

Philip James: Physical activity has a very profound effect on insulin sensitivity, with a whole series of cellular changes that are recognized to occur in the post-receptor responsiveness to normal insulin levels, so I would be surprised if we did not invoke at least one of these mechanisms.

Christian Drevon: Physical activity influences almost every organ system in the body in a beneficial way, and we are searching for nontraditional factors that would go beyond insulin sensitivity and glucose transport, etc. We have found 20 new proteins that we think are myokines secreted or coming from muscle, and 16 of them are responsive to physical activity at the mRNA [messenger RNA] level of control.80 A striking feature of most of these myokines is their relationship to the immune system, which links with Ole Petter's idea. So, this is an interesting link between physical activity and the immune system and also mood, which is very responsive to physical activity.

Mats Rudling: We think that a mediator of physical exercise may be growth hormone, because if you give growth hormone, you can mimic the effect of training in terms of not just muscle improvement but also in reducing plasma cholesterol,81 so I think growth hormone induction after training is one of multiple mechanisms.

SLEEP

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Kaare Norum: Why do we sleep and what is going on with the receptor, the synapse, and so on during sleep?

Richard Wurtman: One of the most effective ways of impairing memory is not to sleep the night before. As people age, they tend to have more and more problems with remaining asleep – not so much falling asleep, but staying asleep. So there is at least superficially a major relationship between success in sleeping and cognitive functions.

Raj Kalaria: Last year I chaired a symposium on trying to find out the incidence and prevalence of dementia in low-income and middle-income countries. One perhaps bizarre observation was that if you look along the time zones where day length is the same across the world, then in all those countries where we could at least get some sort of data, dementia prevalence was low. So, allowing the brain to get some rest may impact cognitive function more generally.

SECULAR CHANGES IN AGING

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES

Marshal Folstein: I just wanted to remind the group that aging in some instances means improvement: one psychological function that improves with age is verbal intelligence. That is how all these gray-headed people here can sit together and hope to come up with an idea that is useful! Verbal intelligence increases with age, and the other thing that happens with aging is that there is a marked decrease in depression and anxiety. We talked about cellular measurement and organ measurement, but a collection of brains operating together forms a culture, and that is another unit. Perhaps cultural aging is important and fueled by increasing verbal intelligence with age.

There is also a well-known phenomenon called the Flynn effect: it has been demonstrated quite clearly that IQ is increasing over the decades from before the First World War, so there is a general increase in intelligence in the population. There is a lot of speculation as to where that comes from, but clearly the mental reserve and capacity is increasing. Whether that has to do with nutrition, we just do not know.

Ingmar Skoog: In Gothenburg we have population studies with exactly the same ages being followed for about 30–35 years. If we compare 70-year-olds examined in 1971 and 70-year-olds examined in 2000, it is a completely different population. They perform much better on all the tests, but they have also completely changed their lifestyles. The answer as to whether that has to do with early factors or later factors is that I think it is both! An old person today is completely different from the person of the same age 30 years ago. The aging process is completely different, and the birth cohort effect needs to be considered. Many people compare a 50-year-old with a 75-year-old, but you are actually not only comparing the two ages, you are also comparing people born in 1934 with those born in the 1950s. So, the birth cohort effect is fooling us a lot. I think the birth cohort effect is often much stronger than the aging effect. Therefore, in our studies, we have, like many others, followed the different age groups prospectively, and what you actually find is that if you look cross-sectionally at aging effects, you have a very strong impression that you get more and more stupid the older you get. But if you follow people longitudinally, you actually do not change so much in cognitive function. You get worse in some things, but you get better in others. So, I think in this meeting when we have been talking about aging, we have sometimes been talking about the aging construed 10–15 years ago, when it was completely different. If you think about a young person today, with mobile telephones, Internet, etc., and all this input, then this will have affected their brain and synaptic connections so that the 25-year-olds today probably have brains that work in a completely different way from that of a 25-year-old 50 years ago.

Philip James: You mean there has been all this adaptation in the brain in response to a multiplicity of new, different, secular-based stimuli?

Ingmar Skoog: Yes, but it is also very important to realize that a 70-year-old today is much healthier than 50 years ago. The sexual drive is much stronger in a 70-year-old today than it was 30 years ago, and people are more happily married than you think; there are changes in almost every aspect of life.

Philip James: So what does happiness do to my decline in brain function?

Ingmar Skoog: The most important factor associated with dying in octogenarians was that they were tired of life. That was much a much stronger predictor than having had a myocardial infarction, stroke, or cancer. Half of those who said they were tired of life had died within 5 years compared with 14% in the rest of the population, and this relationship was completely independent of any physical or mental disorder.

Reiden Torp: It would be good to look at the studies on 2,000 monks in the Mediterranean area, where they are on a so-called Spartan restricted diet. A brief report I have seen suggests that they aged from 50 to 104 years with little or no cardiovascular disease and no Alzheimer's disease.

Kay-Tee Khaw: I would echo Ingmar Skoog's analysis: there are now many epidemiological studies, such as the Rotterdam study, which suggest that different birth cohorts have completely different health expectancies. Not only are the cardiovascular risk factors different but also their cognitive function is very different at the same age in different birth cohorts. We see even in the United Kingdom that people who are now 85 have half the prevalence of disability that people of 85 had just 15 years ago. So, there is a fantastically large shift and clear evidence of the impact of environmental factors.

Philip James: Thank you for this marvelous discussion. I look forward to seeing you all in 35 years' time, sharp as a razor with your verbal skills completely intact!

Declaration of interest.  The authors have no relevant interests to declare.

REFERENCES

  1. Top of page
  2. THE RECEPTOR BIOLOGY OF THE BRAIN
  3. SPECIES DIFFERENCES
  4. ALZHEIMER'S DISEASE AND ITS RELATIONSHIP TO DIET
  5. VASCULAR ISSUES
  6. NUTRITIONAL ASPECTS OF PLASTICITY
  7. TELOMERES
  8. OTHER INFLUENCES: STATINS
  9. A LIFESPAN APPROACH TO AGING
  10. LIPIDS AND AGING
  11. NUTRITIONAL INFLUENCES
  12. DHA, DEVELOPMENT, AND BRAIN FUNCTION
  13. CHOLINE AND B VITAMINS
  14. FOLATE AND NEURAL FUNCTION
  15. FOOD FORTIFICATION
  16. CALORIC RESTRICTION
  17. THE IMMUNE SYSTEM
  18. PHYSICAL ACTIVITY
  19. SLEEP
  20. SECULAR CHANGES IN AGING
  21. REFERENCES