• delayed-response;
  • event-related potentials;
  • genes;
  • QTDT;
  • twins


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

Brain-derived neurotrophic factor (BDNF) may play a role in modulating memory function and there is growing evidence that the BDNF V166M polymorphism may influence episodic memory in humans. However, previous association studies examining this polymorphism and working memory are inconsistent. The current study examined this association in a large sample of adolescent twin-pairs and siblings (785 individuals from 439 families). A range of measures (event-related potential, general performance and reaction time) was obtained from a delayed-response working-memory task and total association was examined using the quantitative transmission disequilibrium tests (QTDT) program. Analyses had approximately 93–97% power (α= 0.05) to detect an association accounting for as little as 2% of the variance in the phenotypes examined. Results indicated that the BDNF V166M polymorphism is not associated with variation in working memory in healthy adolescents.

Brain-derived neurotrophic factor (BDNF) belongs to a family of highly conserved polypeptide growth factors and is an important component in the normal development of the central nervous system (CNS) and in modifying CNS structure and function in adults (Bibel & Barde, 2000; Huang & Reichardt 2001). In humans, it may play a role in the development of neurological conditions such as epilepsy (Scharfman 2005) and in psychiatric disorders such as schizophrenia and bipolar disorder (Fanous et al. 2005; Neves-Pereira et al. 2002).

Extensive studies in rats indicate that BDNF is essential for some forms of learning and memory (Yamada et al. 2002) and suggest an involvement in working memory. For instance, Mizuno et al. (2000) reported that infusion of antisense BDNF oligonucleotide (which was associated with a significant reduction of BDNF mRNA and protein levels in the hippocampus) impaired spatial memory formation, retention and/or recall in rats. Furthermore, frontal cortex BDNF levels are reported to correlate negatively with the number of working-memory errors in aged rats, but not young rats (Bimonte et al. 2003), and in Ts65D mice (an animal model of Down syndrome) (Bimonte-Nelson et al. 2003).

A growing number of studies suggest a role for BDNF in the function of human memory. Egan et al. were the first to examine the V166M polymorphism in the 5′ pro-region of the human BDNF protein in relation to human memory and hippocampal function (Egan et al. 2003) – this polymorphism has also been widely studied for a range of neurological and psychiatric conditions (e.g. Hall et al. 2003; Karamohamed et al. 2005). Egan and colleagues showed that relative to Val, the Met allele was associated with a qualitatively different hippocampal response, as assayed with functional magnetic resonance imaging during performance of an N-back working-memory task. More specifically, a robust and reliable hippocampal deactivation is normally produced in these tasks (Meyer-Lindenberg et al. 2001) and was observed in Val/Val individuals, but an abnormal pattern of increased activity was observed in Val/Met individuals. A subsequent study by Hariri et al. (2003) examined hippocampal and cortical function during the encoding and retrieval of novel, complex scenes – a task previously associated with increased hippocampal activation (Stern et al. 1996). They found hippocampal activity to be reduced in Met-allele carriers compared to Val/Val individuals during both encoding and retrieval. In addition to these observations of disrupted hippocampal function, reduced grey matter volumes for the hippocampus and dorsolateral prefrontal cortex (both regions known to subserve working memory (Fuster 1997)), have been reported for Met-allele carriers compared to Val homozygotes (Bueller et al. 2006; Pezawas et al. 2004; Szeszko et al. 2005).

Egan et al. also examined the association between the V166M polymorphism and a range of memory phenotypes (episodic, semantic, working) (Egan et al. 2003). They found an association with episodic memory [as assessed with the WMS but not the CVLT (Table 1)], such that Met/Met individuals performed poorly compared to Val-allele carriers. This was found in controls alone and also in their entire sample, which comprised individuals with schizophrenia, their siblings and controls. However, they found no association between the BDNF genotype and either semantic or working memory.

Table 1.  Studies investigating associations between the BDNF V166M polymorphism and memory phenotypes
StudySample (Age)Memory phenotype (Test/Test Battery)*Finding
  • *

    WMS, Wechsler Memory Scale; CVLT, California Verbal Learning Test; WCST, Wisconsin Card Sorting Test; VPALT, Verbal Paired-Associate Learning Test; BEES, Behavioral Evaluation for Epidemiologic Studies.

  • Based on demographics reported in Toulopoulou et al. (2003)

  • Note: Episodic and semantic memory are both types of declarative memory. Some of the studies listed above (i.e. Echeverria et al. 2005; Hariri et al. 2003; Strauss et al. 2004) describe their memory phenotype as declarative rather than episodic as shown here. Declarative memory requires conscious recall, with episodic memory based on one’s own experiences and semantic memory based on knowledge learned (Lezak 1995). Episodic and semantic memory are both types of long-term memory, while working memory is a type of short-term memory.

Egan et al. (2003)203 with schizophreniaEpisodic (WMS)Met/Met associated with worse scores for episodic memory obtained from WMS, but not CVLT. No association found with semantic or working memory.
 305 siblingsEpisodic (CVLT) 
 133 healthy controlsSemantic (WCST) 
 (18–60 years)Working (WCST) 
Hariri et al. (2003)64 healthy participants (30.9 ± 1.3 years Val/Val, 30.3 ± 1.6 years Val/Met & Met/Met)Episodic (encode/retrieve fMRI paradigm)Met allele associated with an increased number of recognition errors.
Rybakowski et al. (2003)54 with bipolar (18–72 years, mean 46 years)Working (WCST)Met allele associated with significantly worse scores.
Strauss et al. (2004)63 with major depression or dysthymic disorder (18.4 ± 2.5 years)Episodic (WMS)No association found.
 Episodic (VPALT) 
Dempster et al. (2005)92 with schizophrenia 114 healthy relatives (17–85 years)Episodic (WMS)Met allele associated with significantly worse score in relatives.
Tan et al. (2005)108 with schizophreniaEpisodic (WMS)Met/Met associated with significantly worse scores.
Rybakowski et al. (2006)111 with bipolar (18–72 years, mean 43 years)Working (WCST)Met allele associated with significantly worse scores in bipolar patients for three of five subtests. No differences found for schizophrenia or control groups.
 129 with schizophrenia (18–65 years, mean 27 years) 
 92 healthy controls (19–58 years, mean 31 years) 
Echeverria et al. (2005)194 male dentistsEpisodic (BEES)Met allele associated with 1 Poorer episodic memory in males but not females, and 2 Reduced working memory performance in females but not males – although result was in expected direction.
 233 female dental assistants (all with chronic low-level mercury exposure)Working (BEES) 

As shown in Table 1, a number of studies have since found supporting evidence for an association between the V166M polymorphism and measures of episodic memory such that poorer episodic memory is found in Met-allele carriers compared to Val homozygotes (Dempster et al. 2005; Echeverria et al. 2005; Hariri et al. 2003; Tan et al. 2005), although there are some inconsistencies based on sample type (e.g. healthy vs. schizophrenia) and a lack of association was reported by Strauss et al. (Strauss et al. 2004). In contrast to the findings of Egan et al. (Egan et al. 2003), evidence of an association between working memory and the BDNF V166M polymorphism has since been reported (Echeverria et al. 2005; Rybakowski et al. 2003, 2006), although once again there are some inconsistencies.

Overall, the evidence suggests that the human V166M polymorphism may influence episodic memory performance but with fewer studies, and with the largest study reporting a lack of association (Egan et al. 2003), the evidence for an effect on working memory is less convincing. The present study therefore examined the association between the V166M polymorphism and working memory in a large sample, with greater power than previous studies to detect association in healthy individuals. Furthermore, in addition to measures of overall performance and reaction time, P300 and slow-wave event-related potential (ERP) measures of brain function (all obtained from a delayed-response working-memory task) were examined.

It has been suggested that the hippocampus may be an important generator of the P300 (Fushimi et al. 2005; Halgren et al. 1980; Nakajima et al. 1995; Okada et al. 1983). Thus, as the BDNF V166M polymorphism has been associated with hippocampal function during an N-back working-memory task (Egan et al. 2003), it may also be associated with variation in the P300 elicited during a delayed-response working-memory task. Similarly, the V166M polymorphism has been associated with variation in grey matter volume of the dorsolateral prefrontal cortex (Pezawas et al. 2004) and this may be reflected in P300 and slow-wave measures recorded over the prefrontal brain region. Therefore, it was hypothesized that the P300 and the prefrontal slow wave may be associated with the V166M polymorphism. ERP measures have generally shown little association with behavioural working-memory performance (Hansell et al. 2005) and studies have generally found no association between behavioural measures of working memory and the V166M polymorphism in healthy individuals (Egan et al. 2003; Rybakowski et al. 2006). Consequently, it was hypothesized that no association would be found between the behavioural measures of working memory and the BDNF V166M polymorphism.

Materials and Methods

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


BDNF genotypes and working-memory phenotypes were available for 785 adolescents (385 males and 400 females) from 439 families. These families comprised 193 dizygotic (DZ) twin-pair families with no other siblings, 42 families with a DZ twin-pair plus one other sibling, seven families with a DZ twin pair and two further siblings, six families with three non-twin siblings, 36 families with two non-twin siblings, and 155 single-participant families. [Note that for families with monozygotic (MZ) twin pairs, only one co-twin was included in this study.)] BDNF genotypes were also available for both parents for 292 of these families, for the father only in 26 families, for the mother only in 91 families, and for neither parent in 30 families.

Working-memory data were recorded as part of an ongoing study of cognitive function (Wright & Martin, 2004; Wright et al. 2001). Ethics approval for this study was obtained from the Human Research Ethics Committee at the Queensland Institute of Medical Research. Written, informed consent was obtained from all participants as well as from a parent or guardian. Participants were excluded if the parental report indicated a history of head injury, neurological or psychiatric conditions, substance abuse/dependence, and/or medication with significant CNS effects. All participants were instructed to avoid caffeine-containing foods and drinks in the 2 h before their visit. Testing occurred as close as possible to the participants’ 16th birthday and for those in the present study, age ranged from 15.7 to 22.3 years (mean 16.4 years, SD 0.7 years), with 60 individuals aged 15 years at the time of testing, 649 aged 16 years, 47 aged 17 years, 20 aged 18 years, four aged 19 years, two aged 20 years, and three aged 22 years. Participant ancestry and allele frequencies are shown in Table 2. Of the Val homozygotes, 268 were male (mean 16.3 years, SD 0.7, range 15.7–22.3) and 252 were female (mean 16.4 years, SD 0.8, range 15.7–22.1). Similarly, for the Val/Met genotype, 103 were male (mean 16.4 years, SD 0.7, range 15.7–20.4) and 124 were female (mean 16.3 years, SD 0.5, range 15.7–18.7). Of the Met homozygotes, 14 were male (mean 16.2 years, SD 0.1, range 16.0–16.4) and 24 were female (mean 16.2 years, SD 0.2, range 15.9–16.8).

Table 2.  Distribution of genotypes and alleles for the V166M BDNF variant by ancestry
AncestryNumber of genotypes (%)Number of alleles (%)
  1. Northern European ancestors were from Britain, Denmark, Finland, the Netherlands, France, Poland, Germany, Austria and Switzerland. Mediterranean ancestors were from Italy, Greece, Yugoslavia, Portugal and Turkey.

British Isles (≥50%, n = 598)406 (68)165 (28)27 (4)977 (82)219 (18)
Northern Europe (≥50%, n = 99)60 (61)33 (33)6 (6)153 (77)45 (23)
Mediterranean (≥50%, n = 42)25 (60)14 (33)3 (7)64 (76)20 (24)
Unknown (>50%, n = 46)29 (63)15 (33)2 (4)73 (79)19 (21)
Total (n = 785)520 (66)227 (29)38 (5)1267 (81)303 (19)

Working-memory task

Both ERP and behavioural performance measures were obtained from a computerized visuo-spatial delayed-response task, which has been described previously (Hansell et al. 2001). Briefly, ERP and behavioural data were collected while participants completed a task that required them to remember the location of a visual target. During each trial, participants were required to focus on a central fixation dot to reduce eye movement. Two hundred and fifty milliseconds after fixation onset, a single target (checkered dot, 1.5° visual angle) was presented peripherally. Target presentation was brief in memory trials (150 milliseconds), but in sensory control trials, the target remained on-screen until target location was indicated. Target presentation was followed by a 1- or 4-second delay. In 50% of memory and sensory trials a distracting stimulus (identical to the target but differing in location) was presented for 150 milliseconds during the delay period. The timing of the presentation of the distracting stimulus was random within the interval 300–700 milliseconds post target onset. The disappearance of the fixation dot signalled the end of the delay period and was the cue for participants to lift their preferred hand from a touch-sensitive pad and to indicate target location with a rubber-tipped pointer. In total, eight trial-type variations were presented (memory/sensory × distractor presence/absence × delay 1 second/4 seconds).

ERP recording and processing

ERPs were recorded from 15 scalp locations (Fp1, Fp2, Fz, F3, F4, F7, F8, Cz, C3, C4, Pz, P3, P4, O1, O2) using the Electrocap system. However, in this instance only data recorded at prefrontal (Fp1, Fp2) and parietal (Pz, P3, P4) sites was examined because studies in humans and primates have shown enhanced prefrontal and parietal activation during spatial working memory tasks (Batuev et al. 1985; Fuster 2001; Rowe et al. 2000). Impedances were kept below 5 Ω and linked ears served as reference. Eye movements and blinks were monitored through the placement of electrodes on the supra-orbital ridge and the outer canthus of the left eye. The electro-oculogram (EOG), Fp1 and Fp2 were amplified 5000 times and remaining EEG channels 20 000 times by Grass pre-amplifiers, with a band pass of 0.01–100 Hz. ERPs were sampled at 250 Hz from 100 milliseconds before fixation point onset to 200 milliseconds post fixation point offset and monitored on-line. EEG data exceeding 50 μV root mean squared (RMS) were automatically rejected. Eye blink artefacts were removed using a computerized algorithm developed by examining eye blinks during electroencephalogram (EEG) recordings and using those records as a digital template to detect and eliminate similar patterns from the recordings.

Following artefact rejection, trials were averaged separately for each trial type using a pre-target baseline of 350 ms. The acceptance criteria required that EOG/EEG rejections be less than 40% and that behavioural rejections (too slow, too fast, or spatially incorrect) be less than 30%. Data not meeting these criteria were visually inspected and accepted if the waveforms did not show significant drift and appeared stable (i.e. waveforms from the 1-second delay trials were comparable to those collapsed over the 1-second and 4-second delay trials).

ERP phenotypes

Slow-wave average amplitude, P300 average amplitude, and P300 latency were extracted from the electrophysiological data. Slow-wave average amplitudes were computed for the interval 650–1150 milliseconds post target onset (i.e. the last 500 milliseconds of the 1-second delay period). For the present analyses, slow-wave amplitudes recorded during memory trials in which a distracting stimulus was presented were examined. This trial type was chosen because working memory processes may be better reflected in distractor compared to non-distractor trials (Engle et al. 1999). The mean number of trials averaged for each individual was 62.9 (SD 16.4, range 11–94). Note that recording the slow wave over delay periods resulted in longer trials than is typical for ERP studies, leading to an increased possibility of trial rejection because of eye blinks and other artefacts.

P300 data were collapsed over all trial types because the delay task was not designed to differentiate P300 measures by trial type, and visual inspection of waveforms and preliminary analyses indicated no P300 amplitude or latency differences for trial type (Wright et al. 2002). P300 average amplitude was examined for the interval 150 to 450 milliseconds (post target onset). The mean number of trials per individual was 277.0 (SD 57.8, range 59–376) for P300 average amplitude and 277.7 (SD 57.3, range 59–376) for P300 latency. A detailed description of P300 latency detection has been previously published (Wright et al. 2002).

Preliminary analyses showed high correlations between ERPs recorded at prefrontal sites (0.95–0.97) and between ERPs recorded at parietal sites (0.85–0.93). Consequently, mean prefrontal and mean parietal measures were examined for slow-wave average amplitude, P300 average amplitude and P300 latency. Note that correlations between prefrontal and parietal measures were moderately low, ranging from 0.36 to 0.43 (Table 3).

Table 3.  Means, standard deviations and Pearson correlation coefficients for the Working Memory Phenotypes*
 PrefrontalParietalWins ($)Trials Correct (%)ITime (ms)MTime (ms)
SW (μV)P3 Amp (μV)P3 Lat (ms)SW (μV)P3 Amp (μV)P3 Lat (ms)
  • *

    Slow-Wave Average Amplitude (SW), P300 Amplitude (P3Amp), and P300 Latency (P3Lat)

  • Recorded at prefrontal and parietal sites, Winnings (Wins), Trials Correct (TCorrect), Initiation Time (ITime), and Movement Time (MTime)

  • Analyses are not corrected for twin/sibling relatedness.


Non-ERP phenotypes

Two measures of overall performance (Winnings, Trials Correct) and two measures of reaction time (Initiation Time, Movement Time) were examined. Winnings is the total amount of money won through participation in the delayed-response task – i.e. each correct response was rewarded 2–10 cents, dependent upon accuracy, and each incorrect response (inaccurate, too slow, too fast) was penalized 5 cents. Responses on all trial types contributed to Winnings. Only memory trials were used in the computation of Trials Correct, Initiation Time, and Movement Time. Trials Correct is the percentage of trials within pre-specified accuracy and time constraints (after trials affected by electro-oculo artefacts, as measured by EOG, are removed). Initiation Time is the latency between fixation offset and break of hand contact with the response pad. Movement Time is the latency between break of contact with the response pad and the screen touch time. For a more detailed description of Trials Correct see Luciano et al. (2001) and for Winnings, Initiation Time, and Movement Time, see Luciano et al. (2004).

Zygosity determination and genotyping

Zygosity for same-sex twin pairs was determined using a commercial kit (AmpFISTR Profiler Plus Amplification Kit, Applied Biosystems, Foster City, CA, USA), which was further cross-checked with blood group and other phenotypic data. This method has an overall probability of correct assignment of greater than 99.99% (Nyholt 2006). Using the genotyped data, GRR (Graphical Representation of Relationships; and Relpair were subsequently used to confirm zygosity determination.

Genotyping of single nucleotide polymorphisms (SNP) was performed using primer extension on the Sequenom MassArray system as described previously (James et al. 2004). The V166M polymorphism is identified in the dbSNP public database ( as rs6265. Genotyping error rates were determined by replicate typing of several SNPs on over 3000 individuals and was <0.1% (James et al. 2004). The rs6265 SNP was multiplexed with seven other SNPs for unrelated projects which allowed additional quality checks.


Preliminary analyses

Data were screened for univariate outliers using SPSS 13.0 for Windows (SPSS Inc., 1989–2004). Those data with z-score values greater than ± 3.3 (less than 1% of the dataset) were excluded from all reported analyses (note that results did not differ when analyses were run using a dataset that included outliers). Phenotypes were normally distributed with the exception of Trials Correct, which had moderate negative skewness and required square-root transformation (Tabachnick & Fidell 1989). Furthermore, means, standard deviations and Pearson correlation coefficients were computed.

Allele frequencies were compared for ancestral group using a hybrid approximation to Fisher exact test probabilities for a contingency table using the network algorithm of Mehta and Patel (Clarkson et al. 1993). Families with both parental genotypes available were tested for segregation distortion. Families were classified according to mating genotype (Val/Val×Val/Met, Val/Met×Val/Met, or Val/Met×Met/Met) and offspring genotypes were determined. These were then compared to expected segregation ratios (1:1, 1:2:1, and 1:1 respectively) using a goodness of fit χ2 test. Using merlin (Abecasis et al. 2002), genotype frequencies were examined for deviation from Hardy–Weinberg equilibrium and identity-by-descent (IBD) sharing probabilities were estimated.

Association analyses

Allelic means were evaluated in a family-based approach using the program QTDT (Abecasis et al. 2000a, 2000b). This approach involves maximum-likelihood modelling of the raw data using a variance-components framework that allows for the simultaneous modelling of the means and variances. Analyses tested for population stratification (which can lead to spurious association), locus dominance effects and total association.

To test for population stratification, association effects are partitioned into orthogonal between- and within-family components. Between-family effects (βb) reflect both genuine and spurious association, while within-family effects (βw) reflect only genuine association. Therefore, population stratification is indicated when βbβw. The total association approach is not robust to population stratification, but was chosen in this instance because no evidence of population stratification was found and because it maximizes power by using all available information. Analyses were run as described at, with both sex and age included as covariates.

Genetic power calculations were performed using the package available at (Purcell et al. 2003). P-values for the association analyses can be Bonferroni corrected for a factor of 6 [the number of effective traits as obtained from a Principal Components Analysis (Eigenvalues > 1)]. However, as all association analyses were non-significant before correction for multiple testing, the α-value was left at 0.05 and uncorrected P-values were reported.


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

Preliminary analyses

Means, standard deviations, and Pearson phenotypic correlation coefficients are shown in Table 3. The correlations show that the measures examined reflect a number of largely independent processes occurring during the performance of a working-memory task – i.e. processes underlying prefrontal amplitudes (slow wave and P300), parietal amplitudes (slow wave and P300), P300 latency (prefrontal and parietal), overall non-ERP performance (Winnings and Trials Correct), Initiation Time and Movement Time, (see also Hansell et al. 2005; Luciano et al. 2004; Wright et al. 2002).

Val allele frequencies were similar for all ancestral groups, ranging from 76% to 82% (Table 2). This is consistent with frequencies reported by the International Hapmap Project (The Consortium, 2003) for rs6265, which found a Val allele frequency of 0.825 for a sample of Utah residents with northern and western European ancestry. Allele frequencies were compared for ancestral group and no significant differences were found (P > 0.3).

For families with both parental genotypes, offspring genotypes did not vary significantly from expected segregation ratios (P-values ranged from 0.06 to 0.78). In addition, genotype frequencies did not deviate significantly from Hardy–Weinberg equilibrium in general (Δχ2= 1.19, P= 0.28) or among founders (Δχ2= 0.05, P= 0.82).

Association analyses

No evidence of population stratification was found for any of the phenotypes (Δχ21 range 0.00–1.25, critical value 3.84). Likewise, no indication of locus dominance was found (Δχ21 range 0.00–0.91).

Phenotypic means for the Val/Val, Val/Met and Met/Met genotypes are shown in Fig. 1. No significant findings of association were found between the working memory phenotypes and the BDNF variants (Table 4). Total association maximized the information examined and thus maximized the power to find an association. For these analyses, power ranged from approximately 93% to 97% (α= 0.05) to identify an association accounting for just 2% of the variance.


Figure 1. Means for Val/Val, Val/Met, and Met/Met genotypes for all phenotypes. Slow Wave average amplitude (SW), P300 Amplitude (P3Amp), and P300 Latency (P3Lat) recorded at prefrontal and parietal scalp locations, Winnings (Wins), Trials Correct (TCorrect), Initiation Time (ITime), and Movement Time (MTime) are represented. *Note that phenotypes were standardized to have a mean of zero and a standard deviation of one for ease of comparison.

Download figure to PowerPoint

Table 4.  Tests of association between working-memory phenotypes and BDNF V166M alleles.
 Δχ2 (1df)p
Slow Wave0.040.84
P300 Amplitude0.720.40
P300 Latency0.470.49
Slow Wave3.530.06
P300 Amplitude2.310.13
P300 Latency0.840.36
Trials Correct0.450.50
Initiation Time0.030.86
Movement Time0.030.86


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

BDNF has been previously identified as a potential mediator of memory function (e.g. Yamada & Nabeshima, 2003). There is growing evidence that the V166M polymorphism and episodic memory are associated, with a number of studies reporting poorer episodic memory in Met-allele carriers compared to Val/Val individuals (Dempster et al. 2005; Echeverria et al. 2005; Egan et al. 2003; Hariri et al. 2003; Rybakowski et al. 2003; Tan et al. 2005). However, studies examining the BDNF V166M polymorphism and its association with working memory have not been as consistent in their findings. Studies finding an association in bipolar patients (Rybakowski et al. 2003, 2006) and healthy individuals (albeit with chronic low-level mercury exposure; Echeverria et al. 2005) were offset by studies finding no association in healthy individuals (Egan et al. 2003; Rybakowski et al. 2006), or in individuals with schizophrenia (Egan et al. 2003; Rybakowski et al. 2006).

The present study examined the association between the V166M polymorphism and working memory in healthy adolescents and had considerable power to do so (approximately 93–97% power to detect an association accounting for 2% of the variance of the phenotypes examined). A range of measures obtained from a delayed-response working-memory task, including both event-related potential and behavioural performance measures, were examined. No associations were found.

A finding of no association was consistent with that hypothesized for the non-ERP behavioural measures, which was based upon previous findings (Egan et al. 2003; Rybakowski et al. 2006). However, a lack of association was contrary to that hypothesized for the ERP measures, in particular, for the P300 measures and for slow wave recorded over the prefrontal brain regions.

The hippocampal formation is thought to be a major generator of the P300 (e.g. Okada et al. 1983) and Met-allele carriers are reported to have reduced hippocampal volumes (e.g. Bueller et al. 2006) and abnormal hippocampal function during an N-back working-memory task (Egan et al. 2003). However, the results indicate that any BDNF-related hippocampal deficits did not affect the generation of the P300 component during a delayed-response task. This may be because the P300 has multiple generators (e.g. Halgren et al. 1998) and/or because BDNF-related differences in hippocampal structure and function are not sufficient to affect P300 generation, at least in healthy adolescents. Similarly, any BDNF-related differences in grey matter volumes in the dorsolateral prefrontal cortex (Pezawas et al. 2004) did not appear to influence ERPs recorded over the prefrontal regions.

Consistent findings of an association between the V166M polymorphism and episodic but not working memory may reflect a greater reliance on hippocampal function for episodic versus working memory. A large literature indicates that the hippocampus plays a critical role in episodic memory (Squire et al. 2004). For instance, a recent study by Kramer et al. (2005) examined episodic memory in a sample of neurodegenerative patients and normal older controls, and showed that hippocampal volume predicted recall and recognition memory accuracy, but that frontal lobe, anterior temporal lobe and posterior cortex volumes did not.

In contrast, while the hippocampus is considered critical to the formation of memory networks in general (Fuster 1997), the prefrontal cortex appears to be of vital importance to working-memory function and its role has been widely studied in this context (Fuster 2001). Van Asselen et al. (2006) recently showed the importance of both the prefrontal cortex and the hippocampus to working memory. They examined working memory in stroke patients and healthy controls and found that performance was impaired by damage to the right dorsolateral prefrontal cortex and the bilateral hippocampal formation, in addition to the right posterior parietal cortex. Interestingly, patients with bipolar disorder, for whom an association between the V166M polymorphism and working memory has been found (Rybakowski et al. 2003, 2006), are reported to show impaired prefrontal and hippocampal function during memory-related encoding (Deckersbach et al. 2006). It is also interesting to note that there is some evidence to suggest that the effect of the BDNF V166M genotype may be greater in patient groups than in healthy volunteers (for influence on volume of the hippocampal formation see Szeszko et al. 2005; and for influence on working memory and executive functions see Rybakowski et al. 2006).

A limitation of the current study may be its focus on adolescents. The limited age range provides insight into the genotype/phenotype association in adolescents, but these insights may not be transferable to other age groups. Working memory tasks engage the prefrontal cortex (Fuster 2001) and the human prefrontal cortex is not functionally or structurally mature until early adulthood (Giedd et al. 1999; Hudspeth & Pribram 1992; Levin et al. 1991; Sowell et al. 1999). Furthermore, Webster et al. have shown that in relation to total mRNA, BDNF mRNA levels in the dorsolateral prefrontal cortex are relatively low in adolescents compared to young adults, adults, and aged individuals (Webster et al. 2002). In addition, during adolescence BDNF may be more involved in regulating neuronal morphology and synaptic pruning than in the maintenance of connectivity and synaptic plasticity as in the mature cortex (Webster et al. 2002). A further limitation of the study may be that the delayed-response task is not a standard task used to measure working memory at a behavioural level.

In conclusion, of studies examining the association between the BDNF V166M genotype and memory, this is the first to examine the following:

  • 1
    Phenotypes from a working-memory delayed-response task
  • 2
    Event-related potential measures of memory function, and
  • 3
    Data from a large sample of healthy adolescents.

Furthermore, the study had considerably more power to examine this association in healthy individuals than previous studies. For all measures examined (electrophysiological, general performance and reaction time) no association was found. Therefore, while the loss of efficiency in CNS function associated with the Met allele (Chen et al. 2004) may affect longer-term memory processes underlying episodic memory, in healthy adolescents this loss of efficiency does not appear to affect the shorter-term processes underlying working memory.


  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgments
  • Abecasis, G.R., Cardon, L.R. and Cookson, W.O. (2000a) A general test of association for quantitative traits in nuclear families. Am J Hum Genet, 66, 279292.
  • Abecasis, G.R., Cherny, S.S., Cookson, W.O. and Cardon, L.R. (2002) Merlin-rapid analysis of dense genetic maps using sparse gene flow trees. Nat Genetics, 30, 97101.
  • Abecasis, G.R., Cookson, W.O. and Cardon, L.R. (2000b) Pedigree tests of transmission disequilibrium. Eur J Hum Genet, 8, 545551.
  • Batuev, A.S., Shaffer, V.I. and Orlov, A.A. (1985) Comparative characteristics of unit activity in the prefrontal and parietal areas during delayed performance in monkeys. Behav Brain Res, 16, 5770.
  • Bibel, M. and Barde, Y.-A. (2000) Neurotrophins: key regulators of cell fate and cell shape in the vertebrate nervous system. Genes Dev, 14, 29192937.
  • Bimonte, H.A., Nelson, M.E. and Granholm, A.-C.E. (2003) Age-related deficits as working memory load increases: relationships with growth factors. Neurobiol Aging, 24, 3748.
  • Bimonte-Nelson, H.A., Hunter, C.L., Nelson, M.E. and Granholm, A.-C.E. (2003) Frontal cortex BDNF levels correlate with working memory in an animal model of Down syndrome. Behav Brain Res, 139, 4757.
  • Bueller, J.A., Aftab, M., Sen, S., Gomez-Hassan, D., Burmeister, M. and Zubieta, J.-K. (2006) BDNF Val66Met allele is associated with reduced hippocampal volume in healthy subjects. Biol Psychiatr, 59, 812815.
  • Chen, Z.-Y., Patel, D.D., Sant, G., Meng, C.-X., Teng, K.K., Hempstead, B.L. and Lee, F.S. (2004) Variant brain-derived neurotrophic factor (BDNF) (Met66) alters the intracellular trafficking and activity-dependent secretion of wild-type BDNF in neurosecretory cells and cortical neurons. J Neurosci, 24, 44014411.
  • Clarkson, D.B., Fan, Y. and Joe, H. (1993) A remark on algorithm 643: REXACT: an algorithm for performing Fisher’s exact test in r x c contingency tables. Trans Math Soft, 19, 484488.
  • Deckersbach, T., Dougherty, D.D., Savage, C., McMurrich, S., Fischman, A.J., Nierenberg, A., Sachs, G. and Rauch, S.L. (2006) Impaired recruitment of the dorsolateral prefrontal cortex and hippocampus during encoding in bipolar disorder. Biol Psychiatr, 59, 138146.
  • Dempster, E., Toulopoulou, T., McDonald, C., Bramon, E., Walshe, M., Filbey, F., Wickham, H., Sham, P.C., Murray, R.M. and Collier, D.A. (2005) Association between BDNF val66met genotype and episodic memory. Am J Med Genet B Neuropsychiatr Genet, 134B, 7375.
  • Echeverria, D., Woods, J.S., Heyer, N.J., Rohlman, D.S., Farin, F.M., Bittner, A.C.J., Li, T. and 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., Bertolina, A., Zaitsev, E., Gold, B., Goldman, D., Dean, M., Lu, B. and Weinberger, D.R. (2003) The BDNF val66met polymorphism affects activity-dependent secretion of BDNF and human memory and hippocampal function. Cell, 112, 257269.
  • Engle, R.W., Tuholski, S.W., Laughlin, J.E. and Conway, A.R.A. (1999) Working memory, short-term memory, and general fluid intelligence: A latent-variable approach. J Exp Psychol Gen, 128, 309331.
  • Fanous, A.H., Neale, M.C., Straub, R.E., Webb, B.T., O’Neill, A.F., Walsh, D. and Kendler, K.S. (2005) Clinical features of psychotic disorders and polymorphisms in HT2A, DRD2, DRD4, SLC6A3 (DAT1), and BDNF: a family based association study. Am J Med Genet B Neuropsychiatr Genet, 125B, 6978.
  • Fushimi, M., Matsubuchi, N. and Sekine, A. (2005) Progression of P300 in a patient with bilateral hippocampal lesions. Clin Neurophysiol, 116, 625631.
  • Fuster, J.M. (1997) Network memory. Trends Neurosci, 20, 451459.
  • Fuster, J.M. (2001) The prefrontal cortex - an update: Time is of the essence. Neuron, 30, 319333.
  • Giedd, J.N., Blumenthal, H., Jeffries, N.O., Castellanos, F.X., Liu, H., Zijdenbos, A., Paus, T., Evans, A.C. and Rapoport, J.L. (1999) Brain development during childhood and adolescence: a longitudinal MRI study. Nat Neurosci, 2, 861863.
  • Halgren, E., Marinkovic, K. and Chauvel, P. (1998) Generators of the late cognitive potentials in auditory and visual oddball tasks. Electroencephalogr Clin Neurophsyiol, 106, 156164.
  • Halgren, E., Squires, N.K., Wilson, C.L., Rohrbaugh, J.W., Babb, T.L. and Crandal, P.H. (1980) Endogenous potentials generated in the human hippocampal formation and amygdala by infrequent events. Science, 210, 803805.
  • Hall, D., Dhilla, A., Charalambous, A., Gogos, J.A. and Karayiorgou, M. (2003) Sequence variants of the brain-derived neurotrophic factor (BDNF) gene are strongly associated with obsessive-compulsive disorder. Am J Hum Genet, 73, 370376.
  • Hansell, N.K., Wright, M.J., Geffen, G.M., Geffen, L.B., Smith, G.A. and Martin, N.G. (2001) Genetic influence on ERP slow wave measures of working memory. Behav Genet, 31, 603614.
  • Hansell, N.K., Wright, M.J., Luciano, M., Geffen, G.M., Geffen, L.B. and Martin, N.G. (2005) Genetic covariation between event-related potential (ERP) and behavioral non-ERP measures of working-memory, processing speed, and IQ. Behav Genet, 35, 695706.
  • Hariri, A.R., Goldberg, T.E., Mattay, V.S., Kolachana, B.S., Callicott, J.H., Egan, M.F. and Weinberger, D.R. (2003) Brain-derived neurotrophic factor val66met polymorphism affects human memory-related hippocampal activity and predicts memory performance. J Neurosci, 23, 66906694.
  • Huang, E.J. and Reichardt, L.F. (2001) Neurotrophins: roles in neuronal development and function. Annu Rev Neurosci, 24, 677736.
  • Hudspeth, W.J. and Pribram, K.H. (1992) Psychophysiological indices of cerebral maturation. Int J Psychophysiol, 12, 1929.
  • James, M.R., Hayward, N.K., Dumenil, T., Montgomery, G.W., Martin, N.G. and Duffy, D.L. (2004) Epidermal growth factor (EGF) polymorphism and risk of melanocytic neoplasia. J Invest Dermatol, 123, 760762.
  • Karamohamed, S., Latourelle, J.C., Racette, B.A, et al. (2005) BDNF genetic variants are associated with onset age of familial Parkinson disease: GenePD Study. Neurology, 65, 18231825.
  • Kramer, J.H., Rosen, H.J., Du, A.-T., Schuff, N., Hollnagel, C., Weiner, M.W., Miller, B.L. and Delis, D.C. (2005) Dissociations in hippocampal and frontal contributions to episodic memory performance. Neuropsychology, 19, 799805.
  • Levin, H.S., Culbane, K.A., Hartmann, J., Evankovich, K. and Mattson, A.J. (1991) Developmental changes in performance on tests of purported frontal lobe functioning. Dev Neuropsychol, 7, 377395.
  • Lezak, M.D. (1995) Neuropsychological Assessment, Oxford University Press, New York.
  • Luciano, M., Wright, M.J., Geffen, G.M., Geffen, L.B., Smith, G.A. and Martin, N.G. (2004) Multivariate genetic analysis of cognitive abilities in an adolescent twin sample. Aust J Psychol, 56, 7988.
  • Luciano, M., Wright, M.J., Smith, G.A., Geffen, G.M., Geffen, L.B. and Martin, N.G. (2001) Genetic covariance among measures of information processing speed, working memory, and IQ. Behav Genet, 31, 581592.
  • Meyer-Lindenberg, A., Poline, J.B., Kohn, P.D., Holt, J.L., Egan, M.F., Weinberger, D.R. and Berman, K.F. (2001) Evidence for abnormal cortical functional connectivity during working memory in schizophrenia. Am J Psychiatr, 158, 18091817.
  • Mizuno, M., Yamada, K., Olariu, A., Nawa, H. and Nabeshima, T. (2000) Involvement of brain-derived neurotrophic factor in spatial memory formation and maintenance in a radial arm maze test in rats. J Neurosci, 20, 71167121.
  • Nakajima, Y., Hattori, T. and Mizuno, Y. (1995) Absence of P300 in patients with bilateral hippocampal lesions caused by herpes simplex encephalitis. No To Shinkei, 47, 270275.
  • Neves-Pereira, M., Mundo, E., Muglia, P., King, N., Macciardi, F. and Kennedy, J.L. (2002) The brain-derived neurotrophic factor gene confers susceptibility to bipolar disorder: evidence from a family based association study. Am J Hum Genet, 71, 651655.
  • Nyholt, D. (2006) On the probability of dizygotic twins being concordant for two alleles at multiple polymorphic loci. Twin Res Hum Genet, 9, 194197.
  • Okada, Y.C., Kaufman, L. and Williamson, S.J. (1983) The hippocampal formation as a source of the slow endogenous potentials. Electroencephalogr Clin Neurophysiol, 55, 417426.
  • Pezawas, L., Verchinski, B.A., Mattay, V.S., Callicott, J.H., Kolachana, B.S., Straub, R.E., Egan, M.F., Myer-Lindenberg, A. and Weinberger, D.R. (2004) The brain-derived neurotrophic factor val66met polymorphism and variation in human cortical morphology. J Neurosci, 24, 1009910102.
  • Purcell, S., Cherny, S.S. and Sham, P.C. (2003) Genetic power calculator: design of linkage and association genetic mapping studies of complex traits. Bioinformatics, 19, 149150.
  • Rowe, J.B., Toni, I., Josephs, O., Frackowiak, R.S.J. and Passingham, R.E. (2000) The prefrontal cortex: Response selection or maintenance within working memory. Science, 288, 16561660.
  • Rybakowski, J.K., Borkowska, A., Czerski, P.M., Skibinska, M. and 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. and Hauser, J. (2006) Illness-specific association of val66met BDNF polymorphism with performance on Wisconsin Card Sorting Test in bipolar mood disorder. Mol Psychiatr, 11, 122124.
  • Scharfman, H.E. (2005) Brain-derived neurotrophic factor and epilepsy - a missing link. Epilepsy Curr, 5, 8388.
  • Sowell, E.R., Thompson, P.M., Holmes, C.J., Jernigan, T.L. and Toga, A.W. (1999) In vivo evidence for post-adolescent brain maturation in frontal and striatal regions. Nat Neurosci, 2, 859861.
  • Squire, L.R., Stark, C.E.L. and Clark, R.E. (2004) The medial temporal lobe. Ann Rev Neurosci, 27, 247278.
  • Stern, C.E., Corkin, S., Gonzalez, R.G., Guimaraes, A.R., Baker, J.R., Jennings, P.J., Carr, C.A., Sugiura, R.M., Vedantham, V. and Rosen, B.R. (1996) The hippocampal formation participates in novel picture encoding: evidence from functional magnetic resonance imaging. Proc Natl Acad Sci U S A, 93, 86608665.
  • Strauss, J., Barr, C.L., George, C.J., Ryan, C.M., King, N., Shaikh, S., Kovacs, M. and Kennedy, J.L. (2004) BDNF and COMT polymorphisms. Neuromolecular Med, 5, 181192.
  • Szeszko, P.R., Lipsky, R., Mentschel, C., Robinson, D., Gunduz-Bruce, H., Sevy, S., Ashtari, M., Napolitano, B., Bilder, R.M., Kane, J.M., Goldman, D. and Malhotra, A.K. (2005) Brain-derived neurotrophic factor val66met polymorphism and volume of the hippocampal formation. Mol Psychiatr, 10, 631636.
  • Tabachnick, B.G. and Fidell, L.S. (1989) Using Multivariate Statistics, Harper Collins, New York.
  • Tan, Y.L., Zhou, D.F., Cao, L.Y., Zou, Y.Z., Wu, G.Y. and Zhang, X.Y. (2005) Effect of the BDNF Val66Met genotype on episodic memory in schizophrenia. Schizophr Res, 77, 355356.
  • The International HapMap Consortium. (2003) The international HapMap project. Nature, 426, 789796.
  • Toulopoulou, T., Rabe-Hesketh, S., King, N., Murray, R.M. and Morris, R.G. (2003) Episodic memory in schizophrenic patients and their relatives. Schizophr Res, 63, 261271.
  • Van Asselen, M., Kessels, R.P.C., Neggers, S.F.W., Kappelle, L.J., Frijns, C.J.M. and Postma, A. (2006) Brain areas involved in spatial working memory. Neuropsychologia, 44, 11851194.
  • Webster, M.J., Weickert, C.S., Herman, M. and Kleinman, J.E. (2002) BDNF mRNA expression during postnatal development, maturation and aging of the human prefrontal cortex. Dev Brain Res, 139, 139150.
  • Wright, M., De Geus, E., Ando, J., Luciano, M., Posthuma, D., Ono, Y., Hansell, N., Van Baal, C., Hiraishi, K., Hasegawa, T., Smith, G., Geffen, G., Geffen, L. and Kanba, S. (2001) Genetics of cognition: Outline of a collaborative twin study. Twin Res, 4, 4856.
  • Wright, M.J., Luciano, M., Hansell, N.K., Geffen, G.M., Geffen, L.B. and Martin, N.G. (2002) Genetic sources of covariation among P3(00) and online performance in a delayed response working memory task. Behav Genet, 61, 183202.
  • Wright, M.J. and Martin, N.G. (2004) The Brisbane Adolescent Twin Study: outline of study methods and research projects. Aust J Psychol, 56, 6578.
  • Yamada, K., Mizuno, M. and Nabeshima, T. (2002) Role for brain-derived neurotrophic factor in learning and memory. Life Sci, 70, 735744.
  • Yamada, K. and Nabeshima, T. (2003) Brain-derived neurotrophic factor/TrkB signaling in memory processes. J Pharmacol Sci, 91, 267270.


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

We thank Marlene Grace, Ann Eldridge and Kathleen Moore for twin recruitment and data collection, Megan Campbell for preparation of DNA samples, Troy Dumenil for SNP typing, and the twins and their families for their participation. Phenotype collection was supported by the Australian Research Council (A79600334 (N.M.), A79801419 (G.G., N.M., L.G., M.W.), A79906588 (N.M., G.S., G.G., L.G., M.W.), DP0212016 (N.M., M.W., G.G., L.G.) and the Human Frontier Science Program [RG0154.1998-B (D.B., N.M., J.A.)].