Combining neuropsychology and genetics
in the study of mood disorders
Daniel Smith MRCPsych
Division of Psychiatry, University of Edinburgh
Specialist Registrar in Psychiatry, Southern General Hospital, Glasgow.
daniel.smith@ed.ac.uk
Overview – three reports:
• Recurrent MDD patients versus controls
• ‘pure unipolar’ patients versus bipolar-spectrum
disorder (BSD) patients
• MDD patients with BDNF gene abnormality versus
patients without BDNF gene abnormality
Recurrent early-onset major depressive
disorder (RE-MDD)
• Defined as 2 or more episodes of MDD before age 22
• Is highly morbid (Zubenko et al, 2001)
• Is probably a ‘more genetic’ sub-group of depression
(Maher et al, 2002)
• Carries a high risk of progression to bipolar disorder
(Smith et al, in press)
Depression in young adults
Bipolar-spectrum disorders
Drugs and alcohol
Life events
Clinical characteristics
environmental factors
Early adversity
‘endophenotypes’
Personality traits
Neurocognitive impairment
Genetic risk factors
BDNF
Outline of study
234 consecutive referrals in 12 month recruitment period
90 with current depression and at least one previous episode of depression
87 agreed to participate
Initial assessment:
SCID-1 diagnostic assessment; symptom severity; quality of life; family history; life
events; drug and alcohol use; DSH and suicidal behaviour; blood sample for DNA
After 3 months on treatment:
Symptom severity; Quality of life
Neuropsychology assessment
Personality dimensions assessment
Neuropsychological impairment in mood
disorders
• Affected by clinical sub-type
• Lack of study of homogeneous groups
• Depends on stage of illness
• ‘endophenotype’ theory versus ‘scarring’ effect
• May reflect structural brain abnormalities
Cognitive deficits in depression and bipolar disorder
across mood states
Cognitive
domain
Cognitive task
Attention
Continuous
performance task
(CPT)
Executive
function
Verbal
memory
Euthymic
state
Depressed
state
Manic
state
No
Yes
Yes
Trail-making Test, part
A (TMT-A)
Yes
Yes
No
Digit Symbol
Substitution Test
(DSST)
Yes
Yes
?
Wisconsin Card Sort
Test (WCST)
Yes
?
Yes
Stroop test
Yes
?
Yes
Trail-making Test, part
B (TMT-B)
Yes
Yes
?
California Verbal
learning Test (CVLT)
Yes
Yes
Yes
Brain regions implicated in mood disorders
• Prefrontal cortex
• Dorsolateral:
– Verbal memory
– Attention
– Executive function
• Anterior cingulate:
– Executive function
• Hippocampus
– Verbal memory
Neuropsychological battery in this study
• Assessed for clinical recovery (euthymia) using HRSD (<8)
• Estimate of intellectual functioning:
• National Adult Reading Test (NART)
• Number of years in full-time education
• Block design sub-section score of the WAIS-R
• Verbal memory:
• California Verbal Learning Test (CVLT)
• Executive function:
• Stroop Colour Word Test
• Brixton spatial anticipation test
• Trail-making test A and B
Patients and controls were well matched
MDD patients
(n=63)
Controls
(n=33)
Mean (SD)
Mean (SD)
21.7 (2.25)
22.2 (2.29)
t=-1.22, df=94, p<0.03
43:20
19:14
X2=1.08, df=2, p<0.30
NART IQ
117.5 (3.56)
115.9 (3.69)
t=2.02, df=94, p<0.05
Block design (WAIS)
44.9 (5.96)
45.8 (3.13)
t=-0.85, df=94, p<0.40
Education (years)
16.7 (1.59)
17.3 (1.40)
t=-1.93, df=94, p<0.06
HRSD score
2.6 (1.95)
1.9 (0.93)
t=1.71, df=94, p<0.09
Current medications:
Antidepressants only, n (%)
Mood stabilisers only, n (%)
Both, n (%)
Neither, n (%)
48 (76.2)
8 (12.7)
6 (9.5)
1 (1.4)
0
0
0
0
---------
Characteristic
Age (years)
Gender ratio (F:M)
Significance test and pvalue
Results (1): MDD patients versus controls
Verbal learning
and memory
(CVLT)
MDD patients
(n=63)
Controls
(n=33)
Significance test
Mean (SD)
Mean (SD)
Trial 1
7.0 (1.85)
7.7 (1.79)
t=-1.85, df=94, p<0.07
Trials 1 to 5 total
55.2 (8.42)
60.7 (6.57)
t=-3.30, df=94, p<0.001
Short delay recall
12.5 (2.69)
13.5 (1.91)
t=-1.77, df=94, p<0.08
Long delay recall
12.6 (2.64)
13.9 (1.98)
t=-2.47, df=94, p<0.01
Delayed recognition
minus false positives
14.7 (1.49)
15.2 (1.00)
t=-1.4, df=94, p<0.16
Results (2): MDD patients versus controls
Attention and executive
function
MDD patients
(n=63)
Controls
(n=33)
Significance test
Mean (SD)
Mean (SD)
Stroop colour trial correct
111.6 (0.63)
111.7 (0.47)
t=-0.50, df=94, p<0.62
Stroop colour word trial correct
109.1 (2.58)
110.4 (1.17)
t=-2.77, df=94, p<0.01
Brixton test
6.8 (1.67)
7.4 (1.62)
t=-1.78, df=94, p<0.08
Trail-making Part A (s)
30.3 (8.0)
29.0 (35.3)
t=0.28, df=94, p<0.78
Trail-making Part B (s)
58.6 (16.4)
45.3 (10.9)
t=4.27, df=94, p<0.0001
Conclusion: MDD patients versus controls
• Subtle impairments in:
– Verbal memory
– CVLT trials 1 to 5 total
– Executive function
– Stroop Colour Word errors
– Trail-making Part B
Young people with recurrent depression are at high risk
of progression to bipolar disorder
• 19% of out-patient depressed adolescents develop bipolar disorder 1
• 47% of young adults hospitalised with depression develop bipolar
disorder 2
• Bipolar illnesses tend begin with an episode of depression rather
than mania 3
• Most people with bipolar disorder date the onset of their illness to
adolescence 3
1.
Rao et al., 1995, J. Am. Acad. Child & Adol. Psych., 34, 566-578.
2.
Goldberg et al., 2001, Am. J. Psych., 158, 1265-1270.
3.
Lisj et al., 1994, J. Aff. Disord., 31, 281-294.
Diagnostic criteria for bipolar spectrum disorder (BSD)
A
at least one major depressive episode
B
no spontaneous DSM-IV hypomanic or manic episodes
C
either of the following plus two from D or both plus one from D:
• First degree relative with bipolar disorder
• Antidepressant-induced mania or hypomania
D
if none from C, at least six of the following:
•
•
•
•
•
•
•
•
•
Hyperthymic personality
> 3 depressive episodes
Brief major depressive episodes (< 3 months)
Atypical depressive symptoms
Psychotic major depressive episodes
Early age of onset (< 25)
Postpartum depression
Antidepressant ‘wear-off’ (acute but not prophylactic response)
Lack of response to > 2 antidepressant trials
Ghaemi, S.N., Ko, J.Y., Goodwin, F.K., Cade's Disease and beyond: misdiagnosis, antidepressant use and a
proposed definition for bipolar spectrum disorder. Canadian Journal of Psychiatry, 2002. 47(2): p. 125-134.
Diagnoses (using diagnostic criteria for bipolar
spectrum disorder, BSD)
90
90
80
% subjects
80
% subjects
70
60
70
60
50
BSD
unipolars
50
40
40
30
30
20
20
10
10
0
0
DSM-IV bipolar disorder
DSM-IV MDD
DSM-IV bipolar disorder
DSM-IV MDD
Smith, D.J., Harrison, N., Muir, W., Blackwood, D.H.R., High prevalence of bipolar spectrum disorders in young adults
with recurrent depression: toward a novel diagnostic framework. Journal of Affective Disorders., in press.
‘pure unipolar’ patients versus BSD patients: CVLT, trials 1-5
62
Mean score
CVLT trials 1-5
60
58
56
Bipolar spectrum disorder
(n=21)
'pure unipolar' (n=42)
54
52
Controls (n=33)
DSM-IV MDD (n=63)
50
48
Controls > DSM-IV MDD; p<0.001
Controls > BSD; p<0.001
46
Pure unipolar > BSD; p<0.06
‘pure unipolar’ patients versus BSD patients: Trail-making Test
70
60
seconds
50
Bipolar spectrum disorder
(n=21)
40
'pure unipolar' (n=42)
N.S.
30
Controls (n=33)
20
10
BSD > Controls; p<0.001
BSD > ‘pure unipolar’; p<0.03
0
TMA
TMB
‘pure unipolar’ > Controls; p<0.01
‘pure unipolar’ patients versus BSD patients: Stroop Colour Word Test
4
3.5
errors
3
Bipolar spectrum disorder
(n=21)
2.5
'pure unipolar' (n=42)
2
Controls (n=33)
1.5
1
0.5
BSD > Controls; p<0.002
0
BSD > ‘pure unipolar’; N.S.
Stroop Colour
Stroop Colour-Word
MDD > Controls; N.S.
Conclusion: ‘pure MDD’ patients versus BSD
patients
• BSD patients appear to have more impairment in:
– Verbal memory (CVLT, trials 1-5)
– Executive function (Trail-making test, part B)
• Indirect support for the validity of the proposed diagnostic
criteria for BSD
Intracellular signalling pathways mediating cellular resilience and neuroplasticity
Manji, H. (2003) American Journal of Psychiatry 160 (1) 24.
Brain-derived neurotrophic factor (BDNF)
• Is a neurotrophin found in the neocortex, hippocampus and amygdala 1
• Modulates hippocampal plasticity and hippocampal-dependant memory 1
• Is upregulated by antidepressant therapy 2
• Is associated with bipolar affective disorder 3,4
1.
2.
3.
4.
Lu & Gottschalk (2000) Progress in Brain Research, 128:231-241
Reid & Stewart (2001) British Journal of Psychiatry, 178:299-303
Neves-Pereira et al. (2002) American Journal of Human Genetics, 71:651-655
Sklar et al. (2002) Molecular Psychiatry, 7:579-593
The Val66Met polymorphism of BDNF
– Single nucleotide polymorphism at nucleotide 196 (G/A;
dbSNP number rs6265) produces an amino acid substitution
at codon 66 (Val66Met)
– Approximate frequencies in human controls:
• 68% val/val
• 27% val/met
• 5% met/met
The BDNF val66met polymorphism affects activity-dependant secretion of
BDNF and human memory and hippocampal function.
Egan et al., (2003) Cell, 112;257-269.
– met allele associated with:
• Poorer episodic memory
• Abnormal hippocampal activation on fMRI
• Lower hippocampal intracellular N-acetyl aspartate
– val/met exerts these effects through intracellular trafficking
and activity-dependant secretion of BDNF
Bipolar patients carrying the Met allele of the BDNF polymorphism
perform worse on the Wisconsin Card Sort Test
Rybakowski et al (2003) Bipolar Disorders, 5(6):468-472
Val/Val
(n=44)
Val/Met
(n=9)
Significance test
Mean (SD)
Mean (SD)
Perseverative errors
10.5 (5.0)
17.0 (15.6)
P < 0.03
Non-perseverative errors
9.8 (5.5)
15.9 (15.2)
P < 0.04
Is this polymorphism associated with cognitive
impairment in young adults with recurrent depression?
Genotype frequencies in this sample of MDD patients
(n=40)
• 24 patients were val/val
• 15 were val/met
= 16 with at least one met allele
• 1 was met/met
Comparison groups were well-matched:
Val/Val genotype
N=24
Val/Met or Met/Met
genotype
N=16
Significance test
Mean (SD)
Mean (SD)
Age
21.7 (2.28)
22.1 (1.88)
N.S.
F:M ratio
2:1
3:1
N.S.
MADRS score
2.92 (2.02)
3.06 (1.78)
N.S.
Premorbid IQ:
Number of years of fulltime education
16.8 (1.69)
16.9 (1.50)
N.S.
NART error score
Full-scale IQ
10.5 (2.17)
117.7 (2.71)
10.1 (2.60)
118.1 (3.30)
N.S.
N.S.
Block design WAIS
44.9 (6.87)
46.9 (3.63)
N.S.
Neuropsychological assessment:
Val/Val genotype
N=24
Val/Met or Met/Met
genotype
N=16
Significance test
Mean (SD)
Mean (SD)
CVLT (trials 1-5)
55.3 (8.8)
55.1 (8.1)
N.S.
Brixton raw score
13.0 (4.15)
12.5 (4.00)
N.S.
Trailmaking A
Trailmaking B
B-A
28.8 (6.01)
53.1 (9.05)
24.3 (8.49)
29.6 (7.47)
65.3 (21.1)
37.8 (19.7)
N.S.
P < 0.04
P < 0.04
Stroop C errors
Stroop CW errors
0.21 (0.41)
2.17 (2.06)
0.50 (0.63)
4.56 (3.54)
N.S.
P < 0.02
Conclusions re: BDNF
• Neurocognitive abnormalities associated with this
polymorphism of BDNF are not limited to hippocampal
impairment
• In recurrent, early-onset major depression the val66met BDNF
polymorphism is associated with impaired frontal executive
function but not with hippocampal impairment
• Is this a reflection of a bipolar diathesis in these young
recurrently depressed patients?
Final Summary
• Subtle frontal and hippocampal impairments in euthymic
young adults with recurrent depression compared to controls
– Is this an endophenotypic characteristic?
• Neuropsychological support for the proposed diagnostic criteria
for bipolar-spectrum disorder (BSD)
• Possible to demonstrate that abnormalities of genes implicated
in cognitive function and risk for mood disorder are associated
with cognitive deficits in recovered patients
Acknowledgments
•
Funded by the Kate Hodgson Memorial Fellowship, University of Edinburgh and by the
Health Foundation UK
•
Psychiatric genetics research group in Edinburgh:
– Douglas Blackwood
– Walter Muir
– David Porteous
– Maura Walker
– Margaret van Beck
•
Genotypying lab at the University of Dundee:
– Murray Wilkie
– Gillian Smith
– Roland Wolf
•
Thanks to general practitioners at the University Health Service in Bristo Square, Edinburgh
and community mental health team at Ballendon House, Edinburgh