Enikő Zsoldos1, Nicola Filippini1, Abda Mahmood1, Archana Singh-Manoux2, Mika Kivimäki2, Clare Mackay1, and Klaus P Ebmeier1
1Psychiatry, University of Oxford, Oxford, United Kingdom, 2Epidemiology and Public Health, University College London, London, United Kingdom
Synopsis
Research presented discusses the cumulative effects of Framingham Stroke Risk profile in mid-life on structural brain integrity reduction in older life.Bacground
The repeated occurrence of stressful situations,
leading to ‘allostatic load’, are said to result in cumulative age-dependent
illness [1]. Disease trajectories, such as towards hypertension and stroke are said
to be consequences of allostatic load [2]. Lifetime hypertension predicts brain
structure 28 years later [3]. Multi-measure cardiovascular and stroke risk
scores, such as the Framingham Risk scores [4] predict subsequent cognitive
decline, grey matter volume reduction and white matter changes [5]. The aims of the present analysis are two-fold:
1) to establish whether there is an association between the Framingham Stroke
Risk Profile (FSRP) in mid-life across five phases (P) and reduced structural
brain integrity and 2) to establish whether the predicted effects on brain
structure reduction in older age are the same 5 (P9), 10 (P7), 15 (P5) and 22
(P3) years before as at the time of the scan (P11) [6].
Methods
MRI data from 405 participants of
the Whitehall II imaging sub-study were analysed (age 69.6±5.2 years, range 60
to 83, M:F=321:84). T1 and dMRI scans were acquired as part of the imaging
protocol previously described [7].
Voxelwise GLM analysis was performed
using Randomise in FSL tools [8] to assess the correlation between grey matter
density (GMD) reduction and FSRP at each phase using voxel based morphometry (FSL-VBM)
and fractional anisotropy (FA) and FSRP at each phase using tract-based spatial
statistics (TBSS). In a second analysis, correlations with FSRP at P11
including P9 - P3 in turn as confound regressors were run using both FSL-VBM
and TBSS. Correcting for multiple comparisons, the significance threshold was
set at p < .05, using
threshold-free cluster enhancement (TFCE; [9]).
Lastly, grey matter intensity values
were extracted from the grey matter mask of the third stage of FSL-VBM, a
combined mask of the precuneus and posterior-singulate gyrus (P+PCG) and the
left and right hippocampi separately. Grey matter intensity values of each mask
were correlated with FSRP at each phase, as well as with FSRP at P11
controlling for P9 – P3. FA values were extracted from the frontal white matter
tracts and correlated with FSRP at each phase, as well as with FSRP at P11
controlling for P9 – P3.
Results
FSL-VBM analysis: Significant widespread
correlations were found between GMD reduction and FSRP at P11 – P3. No
significant correlation was found between GMD reduction and FSRP at P11,
controlling for P9. Significant correlations were found between GMD reduction and
FSRP at P11, controlling for P7, P5 and P3. The spread of correlations was
reduced in size, localised to the medial temporal lobes.
TBSS analysis: Significant
widespread correlations were found between FA reduction and FSRP at P11 – P5.
Significant widespread correlations were found between FA reduction and FSRP at
P11, controlling for P9 – P3.
Both FSL-VBM and TBSS findings were
supported by correlations between extracted grey matter intensity / FA values
respectively, and FSRP at each phase, as well as with FSRP at P11 controlling
for P9 – P3.
Discussion
The above findings suggest that 1) a
significant association between FSRP and GMD reduction is present in mid-life
as early as 22 years before the time of the scan. A significant association
between FSRP and FA reduction is also present in mid-life, 15 years before the
time of the scan. 2) The associations between FSRP and GMD reduction are
significantly different in three midlife phases compared to at the time of the
scan in older life. The lack of significant correlation between GMD reduction and
FSRP at P11 controlling for P9 suggests a lack of additional GMD reduction
predicted by FSRP in older age compared to 5 years before the scan. Additional GMD
reduction is predicted by FSRP in older age compared to 10 years before the
scan, in the medial temporal lobes. This additional association in the medial
temporal lobes remains present in relation to FSRP 15 and 22 years before the
scan, due to the plasticity of these areas. In contrast, FSRP in older age does
not predict additional GMD reduction in posterior and cortical areas such as
the P+PCG. 3) The associations between FSRP and FA reduction are significantly
different in four phases compared to at the time of the scan in older age.
These associations become more widespread the further away FSRP is from the
time of the scan in older age.
Conclusions
These findings highlight the importance of
reducing risk factors that contribute to FSRP in midlife, as well as the
differential effect of FSRP on different parts (including grey and white
matter) of the brain.
Acknowledgements
No acknowledgement found.References
[1]
McEwen BS, Stellar E. Stress and the individual. Mechanisms leading to disease.
Arch Intern Med. 1993;153(18):2093-2101.
[2]
Steptoe A, Kivimaki M. Stress and cardiovascular disease. Nat Rev Cardiol.
2012;9(6):360-370.
[3]
Allan CL et al. Lifetime hypertension as a predictor of brain structure in
older adults: cohort study with a 28-year follow-up. Br J Psychiatry 2015; 206(4):308-15.
[4]
D'Agostino RB et al. Stroke risk profile: adjustment for antihypertensive
medication. The Framingham Study. Stroke. 1994;25(1):40-43.
[5]
Allan CL et al. Does the Framingham Stroke Risk Profile predict white-matter
changes in late-life depression? Int Psychogeriatr. 2012;24(4):524-531.
[6]
Marmot M, Brunner E. Cohort Profile: the Whitehall II study. Int J Epidemiol.
2005;34(2):251-256.
[7]
Filippini et al. Study protocol: The Whitehall II imaging sub-study. BMC
Psychiatry. 2014; 14:159.
[8]
Jenkinson M et al. FSL. Neuroimage 2012; 67:782-90.
[9]
Winkler AM et al. Permutation inference for the GLM. Neuroimage 2014;
92:381-397.