• Association;
  • BDNF;
  • cognition;
  • elderly;
  • genes;
  • memory;
  • polymorphism


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

A functional brain-derived neurotrophic factor (BDNF) gene polymorphism (Val66Met) that alters activity-dependent secretion has previously been reported to influence cognitive functioning. A large proportion of these reports suggest that the Met allele, which results in reduced secretion of BDNF, impairs long-term memory as a direct consequence of its influence on hippocampal function but has little influence on working memory. In contrast, other studies have found that the Met allele can also play a protective role in certain neurological conditions and is associated with improved non-verbal reasoning skills in the elderly suggesting effects that appear disease, domain and age specific. We have investigated six haplotype-tagging single nucleotide polymorphisms (SNPs) using a cohort of 722 elderly individuals who have completed cognitive tests that measured the domains of fluid intelligence, processing speed and memory. We found that the presence of the Met allele reduced cognitive performance on all cognitive tests. This reached nominal significance for tests of processing speed (= 0.001), delayed recall (= 0.037) and general intelligence (g) (= 0.008). No association was observed between cognitive tests and any other SNPs once the Val66Met was adjusted for. Our results support initial findings that the Met allele is associated with reduced cognitive functioning. We found no evidence that the Met allele plays a protective role in older non-demented individuals. Magnetic resonance imaging data collected from a subgroup of 61 volunteers showed that the left and right hippocampus were 5.0% and 3.9% smaller, respectively, in those possessing the Met allele, although only a non-significant trend was observed.

The field of cognitive genetics is supporting decades of twin study work that predicted that cognitive ability has a strong genetic basis (Devlin et al. 1997). One gene of particular interest, as a result of its relatively consistent association with long-term memory, is brain-derived neurotrophic factor (BDNF). The BDNF gene codes for a pleotrophic protein that is involved in a variety of neurological functions at different stages of our lives including neurone proliferation and differentiation during brain development, cell survival and synaptic plasticity in the developed brain and oxidative stress in later life (Mattson et al. 2002; Poo 2001; Wang et al. 2006).

Of the 11 publications that have investigated the role of a functional BDNF Val66Met polymorphism and cognitive functioning, nine have reported an association. The first publication reported that the presence of the Met allele not only impaired activity-dependent secretion of BDNF but also reduced delayed episodic (long-term) memory (Egan et al. 2003). The authors found no significant correlation between this polymorphism and other cognitive domains or IQ, suggesting that a domain-specific effect was influencing the hippocampus. The influence of the Met allele on long-term memory was later reproduced by four independent groups (Dempster et al. 2005; Echeverria et al. 2005; Hariri et al. 2003; Tan et al. 2005), and so far, only one group has failed to replicate the association (Strauss et al. 2004). In contrast to the Egan et al. (2003) finding, a number of studies have reported that the Met allele reduces working memory performance (short-term memory) (Echeverria et al. 2005; Rybakowski et al. 2003, 2006) and IQ (Tsai et al. 2004), although these findings have been challenged (Egan et al. 2003; Hansell et al. 2007). A recent investigation of BDNF polymorphisms in Alzheimer’s disease, which is a major cause of cognitive decline in the elderly, found a significant association between a combination of three SNPs that included the Val66Met polymorphism, with heterozygotes conferring susceptibility (Huang et al. 2007). In addition, anatomical magnetic resonance imaging studies have shown that the Met allele significantly reduces hippocampal and cerebral neocortex volume and that these effects appear independent of age and gender (Bueller et al. 2006; Frodl et al. 2007; Pezawas et al. 2004).

In contrast, a recent study that used a cohort of 893 elderly individuals reported that the Met allele is associated with enhanced verbal reasoning ability, which is closely related to IQ (Harris et al. 2006). This study found no association between other measures of cognitive abilities including memory. A protective effect of the Met allele has also been observed in systemic lupus erythematosus and multiple sclerosis patients (Oroszi et al. 2006; Zivadinov et al. 2007).

To test whether the Val66Met polymorphism influences cognitive ability and whether these effects are domain specific, we investigated its influence using a cohort of 722 community-dwelling elderly volunteers. In addition, a further 13 SNPs that formed seven common haplotypes within the BDNF gene were also studied. The Val66Met genotypes were compared against brain region volumes in 61 of the volunteers.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments


DNA was available from 722 elderly community-dwelling Caucasian volunteers from the northwest of England aged between 50 and 85 years, with a mean age of 63 years. Data were available for the potential confounders such as hypertension, diabetes and depression. All volunteers achieved a maximum score on the mini-mental state examination. Details on the recruitment, composition, selective attrition, demographics, health and cognitive tests are described in detail elsewhere (Rabbitt et al. 2004).

Cognitive tests

Fluid intelligence (novel problem solving) was measured using the Heim intelligence test part 1, which required volunteers to answer as many number series, verbal comparisons, logical problems and arithmetical problems as possible in 10 min (Heim 1970). The random letter (RL) test and alphabet-coding task (ACT) were used as measures of processing speed (Savage 1984). The RL test allowed participants 8 min to detect all ‘O’ and ‘I’ letters on four pages of RLs administered as four separate tests of 2 min each. Scores are based on the number of letters spotted and the number omitted. The ACT task requires volunteers to encode 15 different letter sequences using a substitution code as quickly as possible. The test consists of examples and a practice followed by four consecutive trials of 2 min duration each. Memory was assessed using immediate and delayed verbal recall tests and a test of semantic memory. The immediate recall test used a series of 10 six-letter nouns that were flashed one at a time on a screen at 2 second intervals. Volunteers were then asked to write down as many as they remember in any order. Volunteers were given a surprise delayed recall (DR) test 20 min after the session, when they were asked to remember as many of the 10 words as they could. Semantic memory was measured using a series of 15 slides containing five pieces of information: a forename, city of residence, hair colour and occupation of five imaginary individuals that were shown to volunteers for 3 seconds each. Scores were assigned for correct information given to correct person. A common factor was generated from these tests using unrotated principle component analysis to create a score of general intelligence. University of Manchester ethics committee approval and individual written consent to the research was obtained.


HapMap ( was used to select the following 14 SNPs: rs908867, rs12273363, rs2883187, rs2030324, c270t X60202, rs7127507, rs1013402, rs7103411, rs7926362, rs7104207, rs2049045, rs11030104, rs6265 (Val66Met) and rs7124442. All polymerase chain reaction reactions were carried out on PTC-225 Peltier Thermal Cyclers (MJ Research, Waltham, MA, USA) in 384-well microtitre plates using 10 ng of genomic DNA, with a final reaction volume of 10 μl. Five replication samples and two blank controls were used as quality controls for each plate. All laboratory work is performed under the ISO 9001:2000 quality management requirements. Genotyping was performed using Sequenome™ technology.

Measurement of brain volumes

Normalized volumes of the left and right hippocampus, temporal lobe and hemisphere were estimated in 61 elderly subjects using a stereological procedure. For all these subjects, a T1-weighted image pulse sequence with repetition time (TR) of 24 ms, echo time (TE) of 11 ms and flip angle of 30° was acquired on a 1.5 Tesla Philips NT whole body magnetic resonance imaging system. The acquisition provided 160 images comprising contiguous 1.77-mm-thick slices through the brain with a field of view of 23 cm and an acquisition matrix of 256 readings of 128 phase encodings. The images were realigned using the nria2 software, v. 1.0.6 (Brain Behaviour Laboratory, University of Pennsylvania) and linearly interpolated to 296 slices, comprising cubic voxels of size 0.898 mm. The volumes of the left and right hippocampus, temporal lobes and hemispheres were measured using image sections that were oriented perpendicularly to the long axis of the hippocampus. A point-counting procedure was used with structural boundary markers, as previously described (Mackay et al. 1998). These volume estimates were normalized by dividing each subject’s structural volumes by a measure of his or her total cranial volume in order to control for variations in peak brain size in youth.

Statistical analysis

Haplotypes and linkage disequilibrium between markers was calculated using the software HelixTree® (Golden Helix, Inc., Bozeman, MT, USA). Haplotype-tagging SNPs (htSNPs) with a minor allele frequency over 5% were selected using Haploview v.3.32. The minimum pairwise correlation (r2) to select htSNPs was 0.8 (using Tagger software, v. 4.0), which is a widely established threshold. Associations between cognitive ability scores and genotype frequency were determined using least squares multiple regression analysis, which was performed in Stata, v. 8.2 (2001). Test scores were normally distributed. Dominant common factors were identified and scores extracted for the cognitive measures used in the analysis (first eigenvalue 3.10, second eigenvalue 0.68). Data were adjusted for age and gender by covarying their effects. The comorbid medical disorders of hypertension, diabetes and depression were investigated as potential confounders. The majority of cognitive tests were moderately/highly correlated, and this was an attempt to replicate previous findings. Therefore, nominal P values are shown for all results, with permutation analysis performed on significant results only.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

Single nucleotide polymorphism minor allele frequencies are given in Table 1. All genotypes were in Hardy–Weinberg equilibrium. No differences in genotype were observed between the five replication samples on each plate, and the two blank controls showed no amplification. The allele frequencies for the Val66Met (rs6265) were identical to those previously reported in a Caucasian population, with the Met allele having a frequency of 19% (Harris et al. 2006). Linkage disequilibrium (R2 and D′) between the markers is shown in Table 2. The 14 markers formed seven haplotypes, which had frequencies greater than 1% and covered 95% of haplotype diversity (Table 3). The Met allele at codon 66 was found in only one haplotype (haplotype 3), and therefore, analysis of this marker either independently or as a haplotype gave identical results. Six of the 14 markers were htSNPs. Single nucleotide polymorphisms BDNF1, 2 and 5 represented themselves only. Single nucleotide polymorphism BDNF4 captured SNPs BDNF3 and 9. BDNF13 (rs6265) captured BDNF8 and 10–13. Finally, BDNF14 captured BDNF6–7.

Table 1.  Minor allele frequencies and Hardy–Weinberg equilibrium of BDNF SNPs
 MarkerBase pair changeMAFHWE
  1. MAF, minor allele frequency; HWE, Hardy–Weinberg equilibrium.

Table 2.  Linkage disequilibrium R2 and D’ of BDNF markers
DBDNF1 0.160.310.300.720.450.460.
BDNF20.99 0.460.470.110.730.750.260.470.
BDNF30.990.95 0.990.220.600.580.500.960.510.460.490.460.58
BDNF40.960.960.99 0.220.610.590.500.970.510.460.490.460.59
BDNF50.990.990.990.99 0.340.350.
BDNF60.950.970.920.930.99 0.950.340.640.350.310.330.310.97
BDNF70.950.970.910.910.990.96 0.310.610.320.290.310.280.92
BDNF80.960.990.990.990.970.990.91 0.501.000.920.970.920.34
BDNF90.970.990.990.990.970.970.890.99 0.500.920.980.930.34
BDNF100.930.970.970.990.990.990.960.990.99 0.460.490.460.63
BDNF110.980.990.980.970.980.970.890.990.990.98 0.950.990.32
BDNF120.980.990.980.970.970.970.890.990.990.980.99 0.950.34
BDNF130.980.990.980.970.980.970.870.990.990.980.990.99 0.32
Table 3.  Brain-derived neurotrophic factor gene haplotype frequencies
Haplotypers908867rs12273363rs2883187rs2030324C270Trs7127507rs1013402rs7926362rs7104207rs7103411rs2049045rs11030104rs6265rs7124442Haplotype frequency
  1. The seven haplotypes listed represent all those with a frequency greater than 1% and cover 95% of haplotype diversity. The Val66Met polymorphism is rs6265 with the A nucleotide coding the Met allele. Shaded boxes represent the minor alleles.


The strongest associations were observed between the Val66Met polymorphism (rs6265) and tests of processing speed, DR and the general intelligence factor (Table 4). Volunteers homozygous for the Val allele (= 471) had a mean score of 209 (66%) on the RL test (processing speed), compared with 203 (64%) for heterozygous volunteers (= 229, nominal P value 0.072) and 177 (56%) for those homozygous for the Met allele (= 22, nominal P value 0.001). A stepwise decrease in score was observed with all memory tests, with the Met allele being associated with a lower score. However, this only reached statistical significance for the test of DR (nominal P value 0.037). For this test, heterozygous and homozygous Met individuals scored 3% and 13% lower, respectively, compared against homozygous Val individuals. Significant results were also observed for the ACT, which has both a processing speed and memory component. Individuals homozygous for the Val allele had a mean score of 236 (29.6%) compared against 226 (27.8%) for the heterozygous volunteers (nominal P value 0.0001) and 205 (25.7%) for homozygous Met individuals (nominal P value 0.011). Volunteers with the Met allele also scored lower on the test of fluid intelligence, although this failed to reach significance. The general intelligence score was significantly decreased in the presence of one or two Met alleles (nominal P value 0.027 and 0.008, respectively). Allelic permutation analysis (10 000 permutations) was performed on rs6265 for the RL test, ACT test, DR test and our measure of general cognitive ability and remained significant (P value 0.001, 0.001, 0.045 and 0.002, respectively).

Table 4.  Influence of the BDNF htSNPs on the level of cognitive abilities
TesthtSNPMax ScoreMean score HWTMean score HetNominal P value95% CIMean score HMNominal P value95% CI
  1. Comparison of mean scores of Het and HM volunteers against HWT volunteers. ACT, alphabet-coding task; AH1, Heim intelligence tests part 1; CI, confidence interval; g, general intelligence; Het, heterozygous; HM, homozygous mutant; HWT, homozygous wild type; IR, immediate verbal recall; NA, not applicable because of the frequency being lower than 2%; SEM, semantic memory.

AH1rs9088676537.135.90.208−3.16, 0.69NANANA
rs1227336336.537.10.5251.10, 2.1640.50.063−0.19, 7.34
rs203032436.536.90.471−1.15, 2.4937.50.279−0.97, 3.38
C270T36.936.80.847−2.74, 2.25NANANA
rs626537.336.00.168−2.76, 0.4836.50.592−5.58, 3.19
rs712444237.036.50.521−2.08, 1.0537.90.816−2.53, 3.21
RLrs908867316206.7205.90.784−9.17, 6.92NANANA
rs12273363205.2207.60.706−5.45, 8.05212.10.308−7.44, 23.57
rs2030324200.1208.10.0261.04, 16.07209.00.0330.80, 18.64
C270T206.2210.60.485−6.69, 14.10NANANA
rs6265209.2203.60.072−12.72, 0.55177.10.001−48.99, −13.15
rs7124442204.8207.40.638−4.91, 8.00211.50.328−5.90, 17.63
ACTrs908867800231.6230.80.849−10.72, 8.82NANANA
rs12273363228.1235.40.211−2.95, 13.33240.00.290−8.66, 28.92
rs2030324222.2231.80.0191.84, 20.06237.40.0035.57, 27.21
C270T230.8238.00.323−6.22, 18.81NANANA
rs6265236.4222.60.001−22.35, −6.37205.40.011−48.78, −6.42
rs7124442228.3232.90.463−4.87, 10.68235.20.579−10.21, 18.25
IRrs908867107.37.40.501−0.20, 0.40NANANA
rs122733637.37.20.168−0.43,−0.12, 1.03
rs20303247.37.30.860−0.30,−0.22, 0.43
C270T7.37.60.097−0.06, 0.71NANANA
rs62657.47.20.387−0.36, 0.146.860.257−1.06, 0.28
rs71244427.37.20.259−0.38,−0.27, 0.60
DRrs908867105.45.40.987−0.47, 0.48NANANA
rs122733635.45.40.869−0.43, 0.375.80.235−0.36, 1.47
rs20303245.25.40.533−0.30, 0.585.60.201−0.18, 0.87
C270T5.45.40.852−0.67, 0.55NANANA
rs62655.55.20.108−0.72,−2.19, −0.07
rs71244425.45.40.711−0.45, 0.315.70.351−0.36, 1.02
SEMrs908867156.56.50.966−0.52, 0.50NANANA
rs122733636.46.70.390−0.24, 0.626.80.325−0.49, 1.49
rs20303246.46.50.767−0.41, 0.556.60.661−0.44, 0.70
C270T6.56.60.693−0.53, 0.80NANANA
rs62656.66.30.076−0.82,−1.44, 0.87
rs71244426.46.50.931−0.40, 0.436.90.231−0.30, 1.21
grs908867 −−0.18, 0.19NANANA
rs12273363−−0.14,−0.03, 0.61
rs2030324−0.08−0.010.244−0.07,−0.02, 0.36
C270T−−0.18, 0.30NANANA
rs62650.07−0.090.027−0.31, −0.02−0.520.008−0.90, −0.13
rs7124442−0.03−0.010.843−0.16,−0.08, 0.43

The htSNP rs2030324 was also significantly associated with the RL and ACT tests, with those homozygous for the wild-type T allele (= 178) scoring lower than heterozygous (= 375) and homozygous mutant C allele (= 162) individuals. This resulted in a stepwise increase in score of 200 (63.3%), 208 (65.9%) and 209 (66.1%), for the RL test and 222 (27.8%), 232 (29.0%) and 237 (29.7%), respectively, for the ACT test. Individuals who were homozygous for the C allele scored lower on all other cognitive tests compared against homozygous T and heterozygous volunteers, although these failed to reach significance. However, after adjusting for the presence of the Met66 allele, no significant results were observed for rs2030324.

We observed no significant differences between brain volume and the Val66Met polymorphism, although the right and left hippocampus were 3.9% and 5.0% smaller in individuals with at least one copy of the Met allele (Table 5). The right temporal lobe and the right hemisphere were also smaller (1.1% and 2.3%, respectively) in those carrying the Met allele, although the left hemisphere and left temporal lobe were larger (0.5% and 4.2%, respectively).

Table 5.  Influence of the Val66Met polymorphism on brain volumes of homozygous Val and heterozygous individuals
Brain regionMean volume Val/Val, = 43Mean volume Val/Met, = 18% DifferenceP value95% CI
  1. Mean volume units (cm3) have been normalized by dividing structure volume by total cranial volume to control for variations in peak brain size in youth. CI, confidence interval; Lhem, left hemisphere; Lhip, left hippocampus; Ltemp, left temporal lobe; Rhem, right hemisphere; Rhip, right hippocampus; Rtemp, right temporal lobe.

Rhip0.001020.00098−3.90.657−0.00010, 0.00006
Lhip0.001000.00095−5.00.504−0.00009, 0.00004
Rtemp0.029070.02874−1.10.896−0.00195, 0.00171
Ltemp0.028280.02948+4.20.196−0.00063, 0.00299
Rhem0.204760.20002−2.30.286−0.01253, 0.00376
Lhem0.203140.20325+0.50.966−0.00911, 0.00872


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

Since the first association between the Val66Met polymorphism and episodic memory was published (Egan et al. 2003), a steady stream of papers have appeared supporting its function as a regulator of cognitive ability. Using an elderly cohort, we observed that the presence of the Met allele was associated with a decrease in all cognitive test scores. This reached significance for the tests of delayed memory, processing speed and general intelligence and further supports the majority of studies that have found an association between this gene and cognitive performance.

Despite the encouraging associations that have been reported in the literature, there have been some conflicting results. The very first report found that the Met allele was associated with reduced declarative memory, which is dependent on the hippocampus, but not other cognitive domains such as working memory or IQ (Egan et al. 2003). It was suggested that the BDNF polymorphism was acting in a domain-specific manner, which is contrary to previous work that argues that genetic correlations between diverse cognitive abilities are very high and therefore genes have a ‘top–down’ effect on general intelligence (Plomin 1999). The work by Egan and colleagues was supported by several other studies, although the majority of these investigated the Val66Met association with declarative memory only (Dempster et al. 2005; Echeverria et al. 2005; Hariri et al. 2003; Tan et al. 2005). In contrast, three other groups have reported an association with the Met allele and reduced working memory (Echeverria et al. 2005; Rybakowski et al. 2003, 2006). However, the sample sizes of these three studies ranged only from 54 to 233 and a much larger study that utilized 785 individuals failed to find an association with working memory (Hansell et al. 2007). An additional conflicting report was also published from a study using Chinese individuals, which found a significant increase in the Met allele in those with lower IQ (Tsai et al. 2004).

Analysing for the presence or absence of the Met allele in our cohort gives effect sizes of 3.5% for DR and 2% for processing speed. To detect these associations in a Caucasian population, a sample size of approximately 200 for the DR test and 380 for processing speed (which is closely correlated to IQ) would be required. Egan’s study, which used a cohort of 197 normal individuals, had the power to detect the DR effect but not IQ. The Chinese study used an even smaller cohort of 114 but did detect the association with IQ. However, the Met allele has a frequency of 63% in the Chinese population and therefore a smaller sample size would be required to detect its effect. Using a cohort of 722 volunteers, we had 85% power to detect a 2% effect size at a stringent significance level of 0.005 and were therefore able to detect an association between the Met allele and both DR and processing speed. We also observed that the presence of the Met allele was correlated with a non-significant decrease in all other cognitive domains, suggesting that it has a more global influence than has been previously reported although its effects appear to be greater on specific brain regions. The association with our measure of general intelligence, which is a measure of what our diverse tests have in common, supports this observation.

The Met allele has been shown to impair secretion of BDNF, and it has been proposed that this may impact on early and permanent changes in cellular development and plasticity (Bueller et al. 2006; Egan et al. 2003; Frodl et al. 2007; Pezawas et al. 2004). Recently, a large study of 893 elderly Scottish Caucasian individuals has reported that the Met allele improved age-related reasoning skills, which are closely correlated to general intelligence (Harris et al. 2006). The authors suggested that the Val allele may be associated with improved cognitive performance in early life, but in later life it may contribute towards a faster rate of cognitive decline thereby predisposing to cognitive impairment. Our tests of fluid intelligence and processing speed are closely correlated to the reasoning tests, yet for all these tests, the Met allele was associated with reduced performance. The reasons for these contrasting results are unknown, but factors such as dementia and health status, differences in cognitive tests and epistatic interactions may be involved.

Analysis of additional htSNPs found an association between the BDNF SNP4 (rs2030324) and both tests of processing speed. This SNP is located in one of seven non-coding BDNF exons (ENS00000360972), and it has been suggested that it may influence brain region-specific splicing patterns (Liu et al. 2005). Moderate LD exists between this SNP and the Val66Met (R2 = 0.462), which may account for the association. Indeed, no associations between this polymorphism and cognitive tasks were observed once those with the Met allele were removed from the analysis (data not presented). This suggests that Val66Met is likely to be the only polymorphism associated with cognition that we have investigated in this study. No other associations were observed between any remaining htSNPs.

Anatomical magnetic resonance imaging work has reported that the Met allele is associated with a reduction in both hippocampal volume (between 11% and 15%) and grey matter volume in the prefrontal cortex and therefore supports the findings that this allele reduces declarative memory, working memory and IQ (Bueller et al. 2006; Frodl et al. 2007; Pezawas et al. 2004). When we analysed the Val66Met polymorphism against brain volume, we did observe a decrease of right and left hippocampal volume of 3.9% and 5.0%, respectively, in those with the Met allele. However, this failed to reach significance even though our sample size was larger than that used in the original studies.

Multiple testing was an issue in this study given that analysis was performed on six htSNPs and seven cognitive tests for the three possible genotypes. In addition, the volumes of six brain regions were compared in individuals with and without the Met66 allele. However, this was primarily a replication study to investigate whether the Val66Met allele was associated with cognitive ability, and even after permutation analysis, the significant results for this polymorphism remained. It was also encouraging to see a clear additive effect when the genotypes were analysed for all tests of memory, processing speed and most importantly the measure of general intelligence, which further supports the majority of previous findings. Our large sample size of over 700 individuals also adds relevance to the observations. Indeed, lack of statistical power is a major contributor to inconsistencies within the literature and future research needs to address this issue. In addition, BDNF regulates a number of biological mechanisms and is itself under regulation by transcription initiation factors, suggesting a likelihood of epistatic interactions that may determine overall function. Despite some contradictory results regarding the true function of BDNF in cognition, it still remains one of the most interesting genes currently associated with cognitive ability and therefore is deserving of further investigation.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments
  • Bueller, J.A., Aftab, M., Sen, S., Gomez-Hassan, D., Burmeister, M. & Zubieta, J.K. (2006) BDNF Val66Met allele is associated with reduced hippocampal volume in healthy subjects. Biol Psychiatry 59, 812815.
  • Dempster, E., Toulopoulou, T., McDonald, C., Bramon, E., Walshe, M., Filbey, F., Wickham, H., Sham, P.C., Murray, R.M. & Collier, D.A. (2005) Association between BDNF val66 met genotype and episodic memory. Am J Med Genet B Neuropsychiatr Genet 134, 7375.
  • Devlin, B., Daniels, M. & Roeder, K. (1997) The heritability of IQ. Nature 388, 468471.
  • Echeverria, D., Woods, J.S., Heyer, N.J., Rohlman, D.S., Farin, F.M., Bittner, A.C.J., Li, T. & Garabedian, C. (2005) Chronic low level mercury exposure, BDNF polymorphism, and associations with cognitive and motor function. Neurotoxicol Teratol 27, 781796.
  • Egan, M.F., Kojima, M., Callicott, J.H., Goldberg, T.E., Kolachana, B.S., Bertolino, A., Zaitsev, E., Gold, B., Goldman, D., Dean, M., Lu, B. & Weinberger, D.R. (2003) The BDNF val66met polymorphism affects activity-dependent secretion of BDNF and human memory and hippocampal function. Cell 112, 257269.
  • Frodl, T., Schule, C., Schmitt, G., Born, C., Baghai, T., Zill, P., Bottlender, R., Rupprecht, R., Bondy, B., Reiser, M., Moller, H.J. & Meisenzahl, E.M. (2007) Association of the brain-derived neurotrophic factor Val66Met polymorphism with reduced hippocampal volumes in major depression. Arch Gen Psychiatry 64, 410416.
  • Hansell, N.K., James, M.R., Duffy, D.L., Birley, A.J., Luciano, M., Geffen, G.M., Wright, M.J., Montgomery, G.W. & Martin, N.G. (2007) Effect of the BDNF V166M polymorphism on working memory in healthy adolescents. Genes Brain Behav 6, 260268.
  • Hariri, A.R., Goldberg, T.E., Mattay, V.S., Kolachana, B.S., Callicott, J.H., Egan, M.F. & Weinberger, D.R. (2003) Brain-derived neurotrophic factor val66met polymorphism affects human memory-related hippocampal activity and predicts memory performance. J Neurosci 23, 66906694.
  • Harris, S.E., Fox, H., Wright, A.F., Hayward, C., Starr, J.M., Whalley, L.J. & Deary, I.J. (2006) The brain-derived neurotrophic factor Val66Met polymorphism is associated with age-related change in reasoning skills. Mol Psychiatry 11, 505513.
  • Heim, A.W. (1970) AH4 Group Test of General Intelligence. NFER-Nelson, Windsor, UK.
  • Huang, R., Huang, J., Cathcart, H., Smith, S. & Poduslo, S.E. (2007) Genetic variants in brain-derived neurotrophic factor associated with Alzheimer’s disease. J Med Genet 44, 66.
  • Liu, Q.R., Walther, D., Drgon, T., Polesskaya, O., Lesnick, T.G., Strain, K.J., De Andrade, M., Bower, J.H., Maraganore, D.M. & Uhl, G.R. (2005) Human brain derived neurotrophic factor (BDNF) genes, splicing patterns, and assessments of associations with substance abuse and Parkinson’s disease. Am J Med Genet B Neuropsychiatr Genet 134, 93103.
  • Mackay, C.E., Roberts, N., Mayes, A.R., Downes, J.J., Foster, J.K., Mann, D. & Edwards, R.H.T. (1998) An exploratory study of the relationship between face recognition memory and the volume of medial temporal lobe structures in healthy young males. Behav Neurol 11, 320.
  • Mattson, M.P., Chan, S.L. & Duan, W. (2002) Modification of brain aging and neurodegenerative disorders by genes, diet, and behavior. Physiol Rev 82, 637672.
  • Oroszi, G., Lapteva, L., Davis, E., Yarboro, C.H., Weickert, T., Roebuck, T., Bleiberg, J., Rosenstein, D., Pao, M., Lipsky, P.E., Lipsky, R.H. & Illei, G.G. (2006) The Met66 allele of the functional Val66Met polymorphism in the brain-derived neurotrophic factor gene confers protection against neurocognitive dysfunction in systemic lupus erythematosus (SLE). Ann Rheum Dis 10, 13301335.
  • Pezawas, L., Verchinski, B.A., Mattay, V.S., Callicott, J.H., Kolachana, B.S., Straub, R.E., Egan, M.F., Meyer-Lindenberg, A. & Weinberger, D.R. (2004) The brain-derived neurotrophic factor val66met polymorphism and variation in human cortical morphology. J Neurosci 24, 1009910102.
  • Plomin, R. (1999) Genetics and general cognitive ability. Nature 402, C25C29.
  • Poo, M.M. (2001) Neurotrophins as synaptic modulators. Nat Rev Neurosci 2, 2432.
  • Rabbitt, P., Diggle, P., Horan, M., Pendleton, N., Bent, N. & Abson, V. (2004) The University of Manchester longitudinal study of normal healthy old age 1983 through 2003. Aging Neuropsychol Cogn 11, 245279.
  • Rybakowski, J.K., Borkowska, A., Czerski, P.M., Skibinska, M. & Hauser, J. (2003) Polymorphism of the brain-derived neurotrophic factor gene and performance on a cognitive prefrontal test in bipolar patients. Bipolar Disord 5, 468472.
  • Rybakowski, J.K., Borkowska, A., Skibinska, M., Szczepankiewicz, A., Kapelski, P., Leszczynska-Rodziewicz, A., Czerski, P.M. & Hauser, J. (2006) Prefrontal cognition in schizophrenia and bipolar illness in relation to Val66Met polymorphism of the brain-derived neurotrophic factor gene. Psychiatry Clin Neurosci 60, 7076.
  • Savage, R.D. (1984) Alphabet Coding Task-15. Murdock University, Perth, W. Australia.
  • Strauss, J., Barr, C.L., George, C.J., Ryan, C.M., King, N., Shaikh, S., Kovacs, M. & Kennedy, J.L. (2004) BDNF and COMT polymorphisms: relation to memory phenotypes in young adults with childhood-onset mood disorder. Neuromolecular Med 5, 181192.
  • Tan, Y.L., Zhou, D.F., Cao, L.Y., Zou, Y.Z., Wu, G.Y. & Zhang, X.Y. (2005) Effect of the BDNF Val66Met genotype on episodic memory in schizophrenia. Schizophr Res 77, 355356.
  • Tsai, S.J., Hong, C.J., Yu, Y.W. & Chen, T.J. (2004) Association study of a brain-derived neurotrophic factor (BDNF) Val66Met polymorphism and personality trait and intelligence in healthy young females. Neuropsychobiology 49, 1316.
  • Wang, H., Yuan, G., Prabhakar, N.R., Boswell, M. & Katz, D.M. (2006) Secretion of brain-derived neurotrophic factor from PC12 cells in response to oxidative stress requires autocrine dopamine signaling. J Neurochem 96, 694705.
  • Zivadinov, R., Weinstock-Guttman, B., Benedict, R., Tamano-Blanco, M., Hussein, S., Abdelrahman, N., Durfee, J. & Ramanathan, M. (2007) Preservation of gray matter volume in multiple sclerosis patients with the met allele of the rs6265 (Val66Met) SNP of brain derived neurotrophic factor (BDNF). Hum Mol Genet 16, 26592668.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments

The authors thank Research into Ageing, for funding of blood collection and DNA extraction, and Capes Foundation, Brazilian agency for support and evaluation of postgraduate education.