Vitamins are essential micronutrients obtained, at least in part, from dietary sources. These compounds are necessary for normal cell function, physiological processes, growth, and development. By definition they are not synthesized in the amounts required by an organism and; as a consequence of the differing needs of different organisms and differences in abilities to synthesize specific compounds, vitamins differ between different species. In the case of humans, four fat-soluble vitamins (A, D, E, K) and nine water-soluble vitamins (B1, B2, B3, B5, B6, B7, B9, B12, C) have been identified. Micronutrients are the most widely taken food supplements throughout the developed world with a recent study reporting usage in approximately 30% of a European cohort, with higher and more consistent consumption by women (Li et al. 2010). These findings are supported by studies in the United States demonstrating that approximately a third of the adult population had recently consumed vitamin/mineral supplements for their purported health benefits (Radimer et al. 2004; Rock 2007). Among this broad class of supplements, multi-vitamin products are consumed most frequently (Timbo et al. 2006).
Humans lost the ability to synthesize these vitamins in sufficient quantities several hundred million years ago. It has been suggested that this was because the supply of these compounds within the food available made it disadvantageous to continue to synthesize them (Pauling 1970). For example, despite being synthesized by all but a few species of animals, the loss of the ability to synthesize vitamin C (ascorbate) by humans has been proposed to have beneficial effects in terms of the energetic and cellular cost, as well as the oxidative cost associated with synthesizing vitamin C, which in turn leads to a cost in terms of endogenous antioxidant requirements (Banhegyi et al. 1997). It has also been suggested that not only has vitamin synthesis been the subject of evolutionary pressure but that vitamins have also shaped aspects of human evolution (Milton 2000). For instance, the loss of the ability to synthesize vitamin C is purported to have led to an increase in free radical–induced mutations and a sub-sequent acceleration of primate evolution (Challem, 1997). Many of the “diseases of civilization,” such as diabetes, obesity, and cardiovascular disease, may be predicated on the shift away from our evolutionarily determined, largely herbivorous diet, to the high-energy, highly digestible, micronutrient-depleted diet of modern humans in west-ernized societies (Benzie 2003; Milton 2000). As a single example of this, reference to the rate of endogenous synthesis of vitamin C in other mammals, the diet of gorillas (Pauling 1970), and the vitamin constituents of what might be assumed to be a typical pre-agriculture diet (Benzie 2003; Eaton and Konner 1997; Pauling 1970), all suggest that our consumption of vitamin C should be at least ten times the recommended daily allowance espoused by most authorities (e.g., 60 mg in the European Union).
Several indirect strands of evidence that exist to support the efficacy of supple-mentation with vitamins in improving cognition and/or mood will be outlined in the following post, as well as evidence from randomized controlled trials. A large proportion of research in this area has been carried out with elderly cohorts suffering defined cognitive decrements and dementia, which will be covered in a future post. However, it is plausible that interventions of this nature may result in very different effects when studied in a non-elderly population. This review will therefore focus on studies that have utilized healthy, non-elderly samples.
While vitamins are intrinsic to all physiological processes within the body, both fat-soluble and water-soluble vitamins also contribute directly to optimal brain function via a plethora of mechanisms. The two groups of vitamins differ in terms of storage.
Fat-soluble vitamins are stored largely in the liver and adipose tissue, and therefore they persevere for potentially long periods in the body. When taken in excessive quantities, they can also build up to toxic levels. The water-soluble vitamins, on the other hand, are readily excreted from the body in urine, and therefore require more regular and consistent replacement.
FAT-SOLUBLE VITAMINS Vitamin A
Vitamin A (and its bioactive derivative retinoic acid [RA]) plays an important role in embryonic neural development (reviewed by Niederreither and Dolle 2008). Its contribution includes, for instance, neuronal differentiation and neurite outgrowth (for details, see Lane and Bailey 2005; McCaffery et al. 2006). Moreover, recent evidence suggests that vitamin A and its metabolites (RA and all-trans retinoic acid [ATRA]) also have important functions in the adult nervous system (Lane and Bailey 2005; Luo et al. 2009; McCaffery et al. 2006; Olson and Mello 2010). For instance, they have roles in the dopaminergic system (Krezel et al. 1998; Valdenaire et al. 1994, 1998) and hippocampal synaptic plasticity (e.g., long-term potentiation [LTP] and long-term depression [LTD]; Etchamendy et al. 2003; McCaffery et al. 2006; Mey and Mccaffery 2004). Vitamin A has direct effects on cognition, e.g., retinoid signaling influences vocal memory in song birds (Olson and Mello 2010) and in retinoid receptor–deficient mice, learning and memory deficits have been observed alongside changes in LTP and LTD (Chiang et al. 1998; Wietrzych et al. 2005). Animal studies have demonstrated that vitamin A deficiency (VAD) is associated with impaired hippocampal LTP and LTD in mice (Misner et al. 2001), which is reversed by vitamin A administration. In addition, administration of vitamin A or retinoid derivatives ameliorates VAD-mediated learning impairments (e.g., Bonnet et al. 2008; Cocco et al. 2002). High doses, however, can be detrimental (Crandall et al. 2004). Evidence also suggests a role in a number of neurodegenerative and psychiatric disorders, e.g., Parkinson’s disease, Huntington’s disease, schizophrenia, and depression.
Vitamin D
Until comparatively recently, the steroid hormone, vitamin D, was largely associ-ated with its established roles in the regulation of bone mineralization and levels of calcium and phosphorus. More recent evidence has shown that the active form of vitamin D, 1,25(OH)2D3, plays a plethora of roles related to health (Holick 2008), which include a number that are specific to brain function. In vitro/vivo evidence suggests these include a host of neuroprotective and neurotransmission functions, including homeostatic regulation of neuronal calcium, modulation of inducible nitric oxide synthase (iNOS) and upregulation of the endogenous antioxidant glu-tathione. It also upregulates neurotrophin factors, including neurotrophin-3 (NT-3) and glial cell line–derived neurotrophic factor (GDNF), which in turn play a role in synapticity and nerve transmission in the neocortex and hippocampus (Buell and Dawson-Hughes 2008; McCann and Ames 2008). Recent evidence has also con-firmed the presence of vitamin D receptors and the catalytic enzymes involved in the synthesis of 1,25(OH)2D3 throughout the human brain, including cognition-relevant areas (Buell and Dawson-Hughes 2008; McCann and Ames 2008). To date, a wealth of evidence has suggested behavioral decrements in rodents either bred with vita-min D receptor dysfunction or deprived of vitamin D during brain development and beyond (McCann and Ames 2008).
Vitamin E
Vitamin E is composed of the tocotrienols and tocopherols, which are closely related groups of compounds, each of which possesses four analogs, a, 0, y, and S. It is generally accepted that a-tocopherol has the highest biological activity among the compounds (Yang and Wang 2008). On the basis of copious in vitro evidence, it is generally accepted that vitamin E is the brain’s most prevalent lipophilic antioxidant, and in support of this it is notable that it is transported across the blood–brain bar-rier (BBB) by lipoproteins (Goti et al. 2002; Mardones and Rigotti 2004). However, direct in vivo evidence to support this role is scarce, and the molecular mechanisms underlying vitamin E’s physiological roles are still largely undelineated (Brigelius-Flohe 2009). Animal studies indicate a beneficial effect of vitamin E on cognitive function (Fukui et al. 2002; Jhoo et al. 2004; Joseph et al. 1998); and it has been suggested that, other than its potential antioxidant properties, vitamin E might owe its beneficial effects to an indirect role in the inhibition and activation of a raft of essen-tial enzymatic processes, and gene expression (Brigelius-Flohe 2009). Vitamin E deficiency also results in a host of neurological deficits, and supplementation has been suggested as a potential treatment for a variety of neurodegenerative disorders (Ricciarelli et al. 2007).
B Vitamins
The B vitamins play key roles in brain function as co-enzymes and precursors of co-factors in enzymatic processes. In this respect they contribute at some level to all physiological processes within the brain. However, they also have a number of specific roles that might be expected to directly affect aspects of brain function, which in turn may modify behavior, either in the short or long term. For instance, folate (vitamin B9) supplies the methyl group for the conversion of methionine to S-adenosylmethionine (SAMe), and therefore plays a role in the synthesis and integ-rity of DNA and the methylation of proteins, phospholipids, and monoamine and catecholamine neurotransmitters (Mattson and Shea 2003). Similarly, adequate levels of folate and vitamin B12 are required for the remethylation of homocysteine (Hcy), which is a potentially toxic amino acid by-product of one carbon metabolism. Vitamin B6 also plays a key role in this process as a coenzyme of cystathionine synthase and cystathioninelyase, which are required for the metabolism of homocysteine to cysteine (Mattson and Shea 2003). Blocking of the conversion process (e.g., by deficiencies of folate or vitamins B12 and B6) leads to elevated levels of homocysteine, which in turn may contribute to a range of deleterious effects on cellular, hemodynamic, oxidative, and vascular parameters, and ultimately may contribute to a range of neurodegenerative and psychiatric disorders (Reynolds 2006). Vitamin B6 also plays a raft of roles in metabolic processes and is integral to the synthesis of a range of neurotransmitters, including dopamine and serotonin, in its role as a cofactor for aromatic l-amino acid decarboxylase (AADC); an enzyme that catalyzes the decarboxylation of a variety of aromatic l-amino acids. It has also been shown to regulate levels of serotonin (Boadlebiber 1993; Calderon-Guzman et al. 2004).
As well as playing a role in the structure and function of central nervous system, cellular membranes vitamin B1 (thiamine) plays a key role as a co-factor in several enzymatic processes essential for the cerebral metabolism of glucose. At its most extreme, thiamine deficiency, including as a consequence of chronic alcoholism, leads to Wernicke’s encephalopathy; a condition involving selective neuropathological lesions related to dysfunction in neuronal metabolic pathways (Ba 2008).
Vitamin C
Vitamin C (ascorbate) is transported into the brain against a steep concentration gradient and accumulates in neuron-rich areas such as the hippocampus, cortex, and cerebellum at high concentrations (Mefford et al. 1981; Mun et al. 2006). This suggests that it plays a pivotal role in brain function. Within the central nervous system vitamin C plays a plethora of roles, in all of which it functions as a single electron donor. In vitro evidence suggests that vitamin C is a powerful antioxidant. In this respect its roles include reducing oxygen, sulfur, and nitrogen–oxygen radicals generated during normal cellular metabolism, and the recycling of other radicals to their previous forms. An important example of the latter is the reduction of tocopheroxyl radical back to alpha tocopherol (Padayatty et al. 2003). Vitamin C also acts as an essential electron donor for a number of separate enzymes, and, among many other effects, contributes to the synthesis of tyrosine, carnitine, catecholamine neurotransmitters, and peptide hormones (Padayatty et al. 2003). Harrison and May (2009) also elaborate roles for vitamin C in neural maturation, and the neuromodulation of the activity of acetylcholine and the catecholamine neurotransmitters, with resultant direct impacts on behavior in animal models.
Governmentally dictated recommended dietary allowances (RDAs), or similar, of vitamins exist in most developed nations. These RDAs are estimated from the average requirement of individuals within a group/population and the variability in the need for the nutrient among individuals to prevent specific vitamin deficiency diseases such as pellagra, rickets, beri-beri or scurvy, and more general chronic diseases such as osteoporosis and heart disease, in the vast majority (97%–98%) of the population. More recently, most authorities have moved toward adopting dietary reference values (or dietary reference intakes), which incorporate RDAs and simply expand on them to include an “estimated average requirement” and tolerable upper limit.
The U.K. “National Diet and Nutrition Survey” (NDNS) (Ruston et al. 2004) aimed to assess the incidence of biochemical vitamin deficiencies in cross sections of the population aged 19–64 years. The survey reported the results of blood analyte samples for B vitamins and vitamins A, C, D, and E taken from a representative cross section of 1347 respondents. The percent incidence within the population that had abnormally low levels of each vitamin indicative of biochemical depletion/deficiency as defined by a number of established criteria, and which may predispose the individual to specific diseases related to deficiency of the vita-min in question, were presented. This survey showed that 5% of men and 3% of women were biochemically depleted in terms of vitamin C; 2% of men and 4% of women were deficient in vitamin B12; 5% of males and females were margin-ally deficient in red blood cell folate (B9); 10% of males and 11% of females were deficient in vitamin B6, and 66% of both males and females showed marginal or deficient status in vitamin B2 (potentially due to methodological issues). Similarly, averaged across the year, 14% of men and 15% of women had deficient status in vitamin D, with this peaking in the winter months and attenuating during sum-mer. Very similar percentages of prevalence with regard to vitamin C (Schleicher et al. 2009) and vitamin B12 (Evatt et al. 2010) deficiencies have also recently been reported from the United States using National Health and Nutrition Examination Survey (NHANES) data.
However, inconsistencies across studies have arisen as a consequence of differ-ing definitions of deficiency. For instance, the NDNS data show low levels (<1%) of deficiency in terms of serum folate levels when a definition of deficiency of <6.3 nmol/L is applied. This contrasts with serum folate deficiencies of 6% in the French population (sample size = 2102) when a cutoff of 7 nmol/L is applied (Castetbon et al. 2009), and a previous figure of 16% in the U.S. adult population (sample size = 7300) when a cutoff of 6.8 nmol/L was applied. In the latter case, this percentage decreased to 0.5% following the start of mandatory fortification of cereal-grain products with folic acid in 1998 (Pfeiffer et al. 2005). Similar inconsistencies exist in the data for vitamin E status. Ford et al. (2006) note that a wide range of lower cutoff points (from 7 to 28 µmol/L) in serum alpha-tocopherol levels have been employed to indicate vitamin E deficiency. Their own data from 4087 participants would give rates of deficiency ranging from 0.5% of the population at a cutoff point of 11.6 µmol/L to more than 20% of adults in the United States when this is raised to 20 µmol/L. This latter figure would seem to be in better agreement with data showing that the habitual intake of vitamin E is below the “estimated average requirement” in 90% of the adult population of the United States (Ahuja et al. 2004). Similarly, while the NDNS (Ruston et al. 2004) used a figure of 25 nmol/L of 25-hydroxyvitamin D to indicate biochemical deficiency in vitamin D, the current consensus is that a much higher cutoff of <50 nmol/L is indicative of deficiency (Bischoff-Ferrari et al. 2006; Holick 2007), suggesting a much higher prevalence of deficiency in the U.K. population than previously reported.
These figures highlight that for most nutrients information regarding RDAs is either unknown or incomplete, and the recommendations are made on the basis of a number of assumptions and considerations that can lead to large variations in the eventual RDA (Levine et al. 1996; Young 1996). Not only is there a certain amount of inconsistency in definitions of deficiency, but the data also suggest that a sizeable minority of the population may be suffering levels of deficiency that could, at the least, dispose them to a variety of chronic diseases. Given that those in the deficient category represent the tail end of a distribution, and that optimum nutrition must lie some way above the cutoff for insufficiency, it would appear from this that there may be room for improvement in micronutrient status throughout a substantial segment of the population.
The intrinsic importance of vitamins to many aspects of brain function would suggest that a relationship might exist between elements of psychological functioning and vitamin status. This relationship could be seen both in terms of the accrual of physiological damage due to the long-term effects of vitamin status, for instance as a consequence of systemic damage related to oxidative stress or homocysteine levels, or alternatively in terms of effects directly related to current circulating levels. The latter situation may be more amenable to vitamin-related improvements, with the timescale of effects potentially ranging from almost immediately up to the maxi-mum length of time it might take to fully replete physical stores.
The majority of studies in this area have concentrated on cognitive decline and dementia in elderly, at risk or diagnosed cohorts, that may well have suffered an accumulation of systemic damage over many years or decades. However, given the possibility that many healthy adults would be classified as deficient in one or more vitamins, and the fact that the, as yet unidentified, optimum level of vitamins must lie some distance above current definitions of deficiency, it would also seem likely that a relation-ship between vitamin consumption/status and psychological functioning should also be evident in cohorts of non-elderly adults assumed to be cognitively intact or representative of the general population. A relatively small number of epidemiological studies have investigated this question in such cohorts. The studies published after 1994 that employed sample sizes >100 which included, but were not limited to, healthy volunteers under the age of 55 years within their cohort are described in the following.
B VITAMINS AND HOMOCYSTEINE LEVELS
Biochemical Status
The most research attention in this general area has been focused on the relationship between circulating levels of B vitamins and aspects of cognitive function, including cognitive decline, and mood, including depression and anxiety. In terms of healthy, non-elderly cohorts this relationship has been examined with regard to circulating levels of folate, with or without vitamin B12.
Despite a relatively large number of studies exploring this relationship in elderly samples, only three cross-sectional studies (Bjelland et al. 2003; Krieg Jr and Butler 2009; Morris et al. 2003) and two longitudinal studies (Teunissen et al. 2003; Tucker et al. 2005), with follow ups at 6 and 3 years, respectively, met our inclusion criteria. In terms of cognitive performance, Krieg and Butler (2009) found no relationship between folate or B12 and cognition, whereas Teunissen et al. (2003) found a positive relationship between circulating levels of folate and delayed recall but only at baseline. Tucker et al. (2005) found positive relationships between baseline plasma vitamin levels (folate, B6 and B12) and change in constructional praxis as measured by spatial copying at 3 years follow-up. However, only the effects of folate were demonstrated to be independent when effects of the other vitamins were adjusted for.
Morris et al. (2003) explored the role of folate and B12 in depression in those with major depression, those with dysthymia and those with no depression. They found significantly lower levels of red blood cell and serum folate in those with major depression and lower levels of serum folate in those with dysthymia as compared to those with no depression after adjustment for sociodemographic factors. Partial support for these findings comes from Bjelland et al. (2003) who demonstrated an inverse relationship between depression and serum folate, but only in middle-aged women.
Given the intrinsic role that folate and vitamin B12 play in the metabolism of homocysteine, which is a potentially toxic amino acid by-product of one carbon metabolism and is implicated in a number of diseases, all of the studies mentioned earlier also included an assessment of blood levels of homocysteine (Bjelland et al. 2003; Krieg Jr and Butler 2009; Morris et al. 2003; Teunissen et al. 2003; Tucker et al. 2005). Four additional studies also investigated the relationship between homo-cysteine and cognitive performance in sizeable cohorts independently of circulating vitamin levels (Elias et al. 2005, 2008; Schafer 2005; Wright et al. 2004). Six of the studies assessing this relationship reported that higher homocysteine levels were associated in some way with aspects of poorer cognitive function or cognitive decline (Elias et al. 2005, 2008; Schafer 2005; Teunissen et al. 2003; Tucker et al. 2005; Wright et al. 2004), but two of these were only in a subset of elderly participants (Elias et al. 2005; Wright et al. 2004). These effects were also shown to be worse in apolipoprotein E-? 4 carriers in two (Elias et al. 2008; Schafer 2005) out of three studies (Wright et al. 2004). The only other study of cognition and homocysteine found a positive association between homocysteine and digit learn-ing in 20–39 year olds, with no effect in 40–59 year olds (Krieg and Butler 2009). In terms of mood, Morris et al. (2003) found no relationship between homocysteine and depression/depressive symptoms, whereas Bjelland et al. (2003) demonstrated a positive relationship between homocysteine and depression.
Dietary Intake
Although there are several cross-sectional studies that include an examination of the associations between dietary intake of B vitamins and cognitive performance or mood in elderly populations, reference to the dietary intake studies shows that only four studies met our inclusion criteria for non-elderly participants. In a large cross-sectional study, Bryan and Calvaresi (2004) found that intake of vitamins B6, and B12, as assessed by a food frequency questionnaire, were positively associated with subjective perceptions of memory in men; but B12 and folate were negatively associated with subjective memory in women. However, no objective measures of cognitive performance were employed in the study. In an earlier study, using objective measures of cognition, Bryan et al. (2002) found that folate was negatively associated with speed of processing, but positively associated with verbal fluency in a younger group (20–30 years) only. Folate, B12, and B6 were all positively associated with recall in the younger group, and B6 was positively associated with short-delay recall in the younger group and with long-delay recall in an older group (65–92). No age-specific effects were found in the middle group (45–55). Tucker et al. (2005) found that dietary levels of folate, B12, and B6 were positively associated with spatial copying, and that dietary folate was significantly positively correlated with verbal fluency at 3 years. These effects of dietary folate were inde-pendent of the effects of homocysteine or B12 or B6.
Turning to mood measures, one prospective study (Mishra et al. 2009) examined associations between long-term B vitamin intake and subjective mental health (general health questionnaire [GHQ]-28) in a group (N = 636) of 53 year old women, who had been followed-up since birth. The authors found that low B12 intake at 53 years of age was associated with increased psychological distress, and that those with low, as opposed to high, B12 intake throughout adulthood had poorer subjective perceptions of their overall mental health as assessed by the GHQ-28. However, Bryan and Calvaresi (2004) found that higher dietary levels of folate and B6 were related to higher perceived stress levels in women aged 39–65 years. No effects were shown in men. Finally, Bryan et al. (2002) found no effect of folate, B6, or B12 on depression or mood.
Very few studies have assessed the circulating levels of analytes for vitamins A, C, D, and E in non-elderly cohorts. Only three cross-sectional studies (Lee et al. 2009; McGrath et al. 2007; Pan et al. 2009) met our crite-ria, all had sizeable sample sizes and examined the relationship between levels of vitamin D (25(OH)D) and cognitive function and/or mood. Lee et al. (2009) reported a beneficial relationship between vitamin D levels and performance, with low levels of 25(OH)D being associated with reduced digit symbol substi-tution task (DSST) performance in 40–79 year olds. However, McGrath et al. (2007) reported no such effect, with a negative association between vitamin D and learning and memory in their oldest group of participants (60–90 year olds). Perhaps surprisingly no association of 25(OH)D and mood has been established, with both studies within our sample that examined this showing no relationship (Lee et al. 2009; Pan et al. 2009).
Dietary Intake
In terms of dietary intake of vitamins A, C, D, or E and its effects on cognitive function and mood, no studies met our inclusion criteria.
Recent randomized controlled trials examining the effect of vitamin supplementation on cognitive function and mood in healthy non-elderly cohorts. Studies were included that were published after 1994, were undertaken in a non-elderly adult population, had a mini-mum of 30 participants per treatment arm, and employed appropriate placebo and double-blind methodology. Findings are summarized in the following.
Three trials of B vitamin supplementation were identified that conformed to the inclusion criteria. Improvements in psychological functioning were seen in all studies (Benton et al. 1997; Bryan et al. 2002; Durga et al. 2007). Benton et al. (1997) administered vitamin B1 (50 mg thiamine) to young females for 2 months and found improvements in total profile of mood states (POMS) score, and faster decision times on 2, 4, and 8 choice reaction time tasks. Two studies also demonstrated improved function following folic acid alone, with Durga et al. (2007), in the largest (N = 818) and longest (3 years) study reported here, demonstrating improvements across cognitive domains in a cohort of 50–70 year old participants with raised homocysteine levels at the outset. Bryan et al. (2002) also found improvements on a single task (Rey auditory verbal learning test recognition list B) from within a large selection of tasks, but only in an older subsection of their female cohorts that were administered folate for 5 weeks. They also found decrements on a letter fluency task, in comparison to placebo, in the same group and another group administered B12 alone.
No studies that involved the administration of vitamins A, C, D, or E which met the inclusion criteria were identified.
Five studies reported unambiguous improvements in psychological functioning (Carroll et al. 2000; Haskell et al. 2010; Kennedy et al. 2010, 2011; Schlebusch et al. 2000) across their cohorts and two fur-ther studies described improvements that were seen in subsamples, as a function of gender (Benton et al. 1995a,b). Of the five studies demonstrating effects across the cohorts, four administered a similar B vitamin complex plus vitamin C, calcium, and magnesium. In the first two of these studies (Carroll et al. 2000; Schlebusch et al. 2000), supplementation for ~30 days led to improved well-being (psychological general well-being schedule and GHQ-28, respectively), reduced stress (perceived stress scale [PSS] and Berocca stress index), and anxiety (HADS and Hamilton anxiety rating scale) on validated psychometric measures in non-elderly participants. Kennedy et al. (2010) included computerized cognitive testing and replicated the findings with regard to stress (PSS) and general psychological functioning (GHQ-12), and additionally reported increased vigor (POMS), reduced mental tiredness, and improved serial subtraction task performance in non-elderly males. In a methodologically distinct study, a mobile phone assessment was used in the same cohort as Kennedy et al. (2010) to measure effects before and after a day’s work prior to and 7, 14, 21, and 28 days after treatment (Kennedy et al. 2011). Although there were no effects on cognition, the study demonstrated increased “alert” ratings following multi-vitamin supplementation during evening testing on day 14 and during morn-ing and evening assessments on day 28. Concentration and mental stamina ratings were also improved during evening assessments across all days following vitamin supplementation. Physical stamina was improved in the active group during both assessments across each study day. Similarly, using a computerized assessment of cognitive function, Haskell et al. (2010) demonstrated improved cognitive function-ing and reduced physical tiredness following extended multi-tasking in 220 females aged 25–50 years who took a broad multi-vitamin/mineral for 9 weeks.
In terms of the studies that saw differential effects according to gender, Benton et al. found improved mood (Benton et al. 1995b) and improved attentional process-ing (Benton et al. 1995a) in the females within their young adult cohort, with these effects only becoming apparent following 12 months administration of their high-dose multi-vitamin.
In simple terms, the rationale for the current review was that vitamins are intrinsically involved in every aspect of brain function, and that our modern, micronutrient-poor diets predispose us to consume less than the optimal levels of vitamins; therefore raising the possibility that, firstly, intake/levels of vitamins might be related to cog-nitive function, and, secondly, that supplementation with vitamins might improve psychological functioning.
The epidemiological evidence relating biochemical levels and dietary intake of vitamins to psychological functioning is currently insufficient to make any conclusion on this in this population at present. By far, the most research in this area has concentrated on the B vitamins, most notably folate, and vitamins B6 and B12. In terms of folate, Tucker et al. (2005) found improvements to spatial copying in 50–85 year olds were related to both baseline biochemical and dietary folate. Improvements to verbal fluency were also related to baseline dietary folate intake and these effects were independent of intake of other vitamins and homocysteine levels. Teunissen et al. (2003) also found a positive association between serum folate levels and delayed recall at baseline in 30–80 year olds, but this effect was no longer apparent at the 6 year follow-up and there were no effects on word learning, Stroop, or letter–digit coding. No relationship between serum folate and simple reaction time, digit–symbol substitution, or serial digit learning was observed by Krieg and Butler (2009) in 20–59 year olds, whereas Bryan and Calvaresi (2004) found an inverse relationship between dietary folate and subjective memory in women aged 39–65. Bryan et al. (2002) also found a negative relationship between dietary folate and speed of information processing in 20–92 year olds. However, a positive relationship was shown for the younger group only (20–30 year olds) in terms of verbal fluency and recall.
Tucker et al. (2005) found improvements to spatial copying in 50–85 year olds were related to both baseline biochemical and dietary B6 and B12 levels. Similarly, Bryan and Calvaresi (2004) found a positive relationship with subjective memory for B6 and B12 in men. Bryan et al. (2002) found that B6 was positively correlated with short-delay recall, and B6 and B12 were positively associated with recall in younger group (20–30 year olds). In an older group (65–92 year olds) B6 was positively cor-related with long-delay recall. No specific effects were discovered in a middle age group. Teunissen et al. (2003) found no effects of B12 on word learning, Stroop or letter–digit coding and no relationship between B12 and simple reaction time, digit– symbol substitution or serial digit learning was observed by Krieg and Butler (2009) in 20–59 year olds.
Seven of the studies identified included a measure of the relationship between homocysteine and cognition, of which six showed some evidence for a negative relationship between the two. Teunissen et al. (2003) demonstrated that higher homocysteine levels were related to impaired Stroop and word learning. Tucker et al. (2005) observed that homocysteine was inversely related to spatial copying and word recall. The results of two cross-sectional studies supported this negative relationship and indicated that this was stronger in ApoE4 carriers (Elias et al. 2008; Schafer 2005). Two other cross-sectional studies only demonstrated these effects in elderly sub-groups (Elias et al. 2005; Wright et al. 2004), and one final study observed a positive relationship between homocysteine levels and digit learn-ing in 20–39 year olds.
In terms of mood there were five studies that met our inclusion criteria, of which three showed some element of a positive association, one showed no effects, and one demonstrated a negative relationship between B vitamins and mood. Morris et al. (2003) observed higher serum folate in 15–39 year olds who had never been depressed than those with major depression or dysthymia. Bjelland et al. (2003) also found that serum folate was inversely related to depression score but only in middle aged women (46–49 years), with no effect in men or women aged 70–74 years. In a prospective study with 53 year follow-up, Mishra et al. (2009) found that low dietary B12 intake at age 53 was associated with increased psychological distress. However, Bryan et al. (2002) failed to find any relationship between dietary folate, B6, or B12 and depression and mood in participants aged 20–92 years. Bryan and Calvaresi (2004) also failed to find any relationship between folate, B6, or B12 and depres-sion, anxiety, or self-esteem. However, perceived stress was inversely correlated with dietary folate and B6, but only in women. Only two studies explored the relation-ship between homocysteine and mood with one of these finding a positive correlation between homocysteine levels and depression in 46–74 year olds (Bjelland et al. 2003), and the other finding no relationship in 15–39 year olds (Morris et al. 2003).
The evidence with regard to the other vitamins (A, C, D, and E) is very limited in this population. Two studies explored the relationship between serum 25(OH)D and cognition, with one of these finding a positive relationship between vitamin D and digit symbol substitution performance in 40–79 year old males (Lee et al. 2009) and the other finding an impairment to learning and memory associated with higher levels of vitamin D in an older subset of participants (60–90 years). There were no effects in the younger subset (20–60 years). Surprisingly, no relationship between vitamin D and mood was established. Serum 25(OH)D was not associated with physical activity, mood, or depression in 40–79 year old males (Lee et al. 2009); nor was plasma 25(OH)D related to depression in 50–70 year olds (Pan et al. 2009).
It is difficult to make any firm conclusions on the basis of the findings from epidemiological research. Naturally, any interpretation of epidemiological evidence is also complicated by the inability either to attribute cause and effect or rule out the influence of a plethora of other potential factors that might co-vary with vitamin status, which may not have been identified in the statistical models employed. Examples of the latter may include, for instance, aspects of socio-economic status, education, or healthy living practices. One particular weakness in this area is also the predominant use of elderly cohorts, making any meaningful assessment of the relationship in non-elderly populations very difficult. While the cognitive decline associated with old age may provide a sensitive backdrop for examining the effects of dietary habits in the context of a large part of the lifespan, investigations in these age groups are also complicated by a number of factors. These include age-related changes in the ability to absorb and metabolize vitamins (Wolters et al. 2004), which may render the results of analyte studies less meaningful, and a relationship between age, declin-ing health, and cognitive function (Payette and Shatenstein 2005; Shatenstein et al. 2007), which suggests that the elderly develop atypical diets, and may well predispose the less cognitively able individual to seek and consume a poorer diet.
In comparison to the epidemiological evidence, research investigating the effects of vitamin supplementation in non-elderly cohorts is even sparser. Only three studies of B vitamins and no studies of the other vitamins (A, C, D, and E) were identified that met our inclusion criteria. Of the studies covering B vitamins, all three discovered a positive impact of supplementation on at least one aspect of cognition and/or mood. Benton et al. (1997) observed improvements to decision times and mood fol-lowing 2 month supplementation with thiamine in healthy young females. Folic acid supplementation for 3 years in 50–70 year olds was also shown to improve cognition and to decrease homocysteine at years 1, 2, and 3 (Durga et al. 2007). However, Bryan et al. (2002) found that 5 week supplementation with folate only improved recognition in an older subset (65–92 years) and folate and B12 impaired verbal flu-ency in all participants (20–92 years).
Naturally, while attractive from the point of view of attributing any treatment-related effects, the focus of these studies on individual vitamins lowers the likelihood of targeting a cohort with a specific requirement for increased levels of the vitamin being supplemented, or alternatively raises the possibility of failing to increase the reduced levels of other micronutrients that might coexist in the individual as a consequence of poor general diet. The evidence with regard to multi-vitamins offers some support for this suggestion and shows some promise with regard to supplementation, particularly in non-elderly populations. Interestingly, the multi-vitamins studies included have tended to be con-ducted solely in younger cohorts rather than the broad range used in epidemiological studies. The pattern of results may reflect the intactness of the cohorts in terms of the physiological mechanisms that might be modulated, and this would provide an argument for focusing on supplementation/dietary improvements in younger cohorts as a means of preventing cognitive decline rather than attempting to “cure” the effects of sub-optimal nutrition at a stage where it may be too late. The effects seen with multi-vitamins may also relate to the broad nature of the treatments, in that the use of multiple vitamins should be more likely to bolster an individual’s requirement for one or more vitamins. It is interesting to note in this respect that there are marked inter-individual differences in the absorption and excretion of vitamins (Shibata et al. 2005, 2009) as a consequence of a number of factors, including genetic makeup, gen-der, and ethnicity (Caudill 2009; Kauwell et al. 2000). In this respect, it should be noted that RDAs are merely population statistics and of very little use in identifying the required minimum daily intake of a nutrient for any individual.
Interestingly, while the strongest relationship seen in the epidemiological studies reviewed earlier is that between circulating levels of homocysteine and cognitive function, homocysteine levels were only measured in two of the intervention studies. Both studies showed a reduction in homocysteine levels and an improve-ment in cognition following vitamin supplementation. Although neither study correlated the improvements to cognition with the changes in homocysteine levels, Durga et al. (2007) found the greatest improvement to cognition in those who had homo-cysteine levels =12.9 µmol/L at baseline. However, the inter-relationships between B vitamins and homocysteine levels, and the causality of the relationships with brain function, are far from clear (Elias et al. 2006; Krieg and Butler 2009), and it remains a possibility that homocysteine levels represent an epiphenomenon reflecting other, as yet un-delineated factors. There is also the possibility that high homocysteine levels do not represent a problem until older age and may also be in part related to ApoE-4 status.
Given that the optimum level of vitamin consumption must reside some way above deficiency levels, it is possible that a large subsection of the general, non-elderly population must have less than optimal micronutrient status. More research might therefore be usefully directed toward delineating the optimal levels of vitamins and their relationships with brain function in non-elderly humans. This research should take advantage of the many sensitive computerized measures of cognitive function that are now readily available. The one area where the preponderance of studies has been conducted in younger samples is that assessing the effects of multi-vitamins. The literature here suggests that non-elderly healthy adults might derive benefits from supplementation in terms of psychological functioning. This in itself supports the notion of less than optimal nutritional status in the population, but also suggests that broad multi-vitamin treatments, which might provide different benefits to individuals on the basis of their personal nutritional status, might be more useful than single/several vitamin supplements.
Aucun commentaire:
Enregistrer un commentaire