Cognitive Performance in

1
Cognitive Performance in
Vitamin B12 Deficient Vegan Men with
Intermediate Hyperhomocysteinaemia
Vaughan Bell 1, Zouë Lloyd-Wright 2, Jan Møller 3,
Anne-Mette Hvas 3, Ebba Nexø 3, Virginia Ng 2,
Steven Williams 2, Tim Key 4, Tom A. B. Sanders 2
1 Cardiff University, 2 Kings College London,
3 Aarhus University Hospital, 4 Oxford University
2
Aims

• Previous studies have typically worked with
• Older adults / demented patients
• Inappropriate cognitive tests (e.g. MMSE)
• We targeted low / normal B12 vegans
• Otherwise healthy, with full age range.
• Used cross-sectional / double-blind RCT
  supplementation design.
• Relevant and robust neuropsychological tests.
• Comprehensive structural neuroimaging.

3
Sample
Vegan males, N 138
Three groups No significant differences
between groups on age, mean veganism, NART IQ
4
Homocysteine
Baseline total mean 19.4 µmol/L (22.6)
SD 36.89
SD 13.38
SD 2.54
ANOVA p lt 0.0005, Pearson r -.352, p lt 0.0005
5
Design

• Phase 1 Baseline assessment
• Comparison between B12 groups
• Phase 2 Double blind RCT Four groups
• 5mg/day B12 supplementation
• Tested again after 3 months
• Compared active / placebo

6
Effect of Supplementation
7
Cognitive Tests

• Tests were chosen to be
• Sensitive to sub-clinical and clinical deficits.
• Have known links to functional neuroanatomy.
• Well controlled (computer presented)
• And for semantic and working memory tasks
• Have varying levels of demand.
• Have clear patterns of performance in healthy
  participants.

8
Episodic Memory
One way ANOVA (B12 Group)
Two-way mixed ANOVA (Phase x Treatment)

• Free recall
• No effect of group (p 0.94), phase (p 0.62)
• No phase x treatment interaction (p 0.39)
• Single-probe recognition
• No effect of group (p 0.55), phase (p 0.19)
• No phase x treatment interaction (p 0.22)

9
Semantic Memory
Multi-factorial sentence verification task
(Kounios Holcomb, 1992)
ALL BIRDS ARE CROWS NO VEGETABLES ARE HAMMERS

• Semantic manipulation
• High / low relatedness
• Category / exemplar order
• All / No sentence

10
Semantic Memory
One way ANOVA (B12 Group)
Two-way mixed ANOVA (Phase x Treatment)

• No main effect of group (p 0.37)
• Main effect for phase (p 0.004, practice
  effect)
• No phase x treatment interaction (p 0.90)
• All semantic manipulations had expected effect

11
Working memory
n-back task (Braver et al., 1997)

• Random letters appear on-screen one by one.
• Four conditions 0, 1, 2 and 3-back
• Participants must indicate if the letter
  on-screen matches the letter n-back.
• 33 targets in each condition.
• Prefrontal cortex involvement directly related to
  working memory load.

12
Working memory
Phase 1
13
Working memory
Phase 2
14
Working memory
n-back task (Braver et al., 1997)

• 3 way ANOVA treatment x phase x n-back
• Main effect for n-back (p lt 0.0005)
• No main effect for phase (p 0.092)
• No main effect for treatment (p 0.586)
• No interactions

15
Neuroimaging
18 males with lt 120ng/L for 10 years, brain /
spine scans with 1.5 Telsa structural MRI (T1 /
T2 / FLAIR)

• 50 yr vegan male with dorsal spinal column
  demyelination
• T2 weighted MRI scan of upper thoracic area.

Nakamura and Swanson (2004)
Present study
16
Neuroimaging
No evidence of brain / spinal column
neuropathology when assessed by a consultant
radiologist.
No evidence of clinical signs of B12 deficiency
(e.g. tingling, fatigue, weakness,
confusion)
17
Very Low B12 Participants

• Imaging included 3 participants with very low B12
  levels (8, 33 and 38 ng/L)
• Cognitive tests showed no obvious deficits

18
Possible explanation
Compensation from excellent cerebrovascular
function and high levels of folate
UK omnivore male mean
No main effect (p 0.78), no post-hoc differences
19
Possible explanation

• e.g. low folate largely known to be linked to
  poor cognition in older adults, independent of
  B12, B6, homocysteine levels (Kado et al., 2005).
• Although see Morris et al. (2005).

20
Conclusions

• Low B12 and hyperhomocysteinanemia not
  necessarily associated with
• Cognitive dysfunction
• Neuropathology
• Our sample possibly protected by
• high levels of folate
• excellent cerebrovascular function