Contributions of cardiovascular risk and smoking to chronic obstructive pulmonary disease (COPD)-related changes in brain structure and function

Abstract

Background

Brain damage and cardiovascular disease are extra-pulmonary manifestations of chronic obstructive pulmonary disease (COPD). Cardiovascular risk factors and smoking are contributors to neurodegeneration. This study investigates whether there is a specific, COPD-related deterioration in brain structure and function independent of cardiovascular risk factors and smoking.

Materials and methods

Neuroimaging and clinical markers of brain structure (micro- and macro-) and function (cognitive function and mood) were compared between 27 stable COPD patients (age: 63.0±9.1 years, 59.3% male, forced expiratory volume in 1 second [FEV₁]: 58.1±18.0% pred.) and 23 non-COPD controls with >10 pack years smoking (age: 66.6±7.5 years, 52.2% male, FEV₁: 100.6±19.1% pred.). Clinical relationships and group interactions with brain structure were also tested. All statistical analyses included correction for cardiovascular risk factors, smoking, and aortic stiffness.

Results

COPD patients had significantly worse cognitive function (p=0.011), lower mood (p=0.046), and greater gray matter atrophy (p=0.020). In COPD patients, lower mood was associated with markers of white matter (WM) microstructural damage (p<0.001), and lower lung function (FEV₁/forced vital capacity and FEV₁) with markers of both WM macro (p=0.047) and microstructural damage (p=0.028).

Conclusion

COPD is associated with both structural (gray matter atrophy) and functional (worse cognitive function and mood) brain changes that cannot be explained by measures of cardiovascular risk, aortic stiffness, or smoking history alone. These results have important implications to guide the development of new interventions to prevent or delay progression of neuropsychiatric comorbidities in COPD. Relationships found between mood and microstructural abnormalities suggest that in COPD, anxiety, and depression may occur secondary to WM damage. This could be used to better understand disabling symptoms such as breathlessness, improve health status, and reduce hospital admissions.

Keywords:

• chronic lung disease
• cigarette smoke
• cognition
• depression
• MRI
• neuroimaging

Supplementary materials

Inclusion/Exclusion criteria

Table S1 Inclusion and exclusion criteria

Clinical measures

Hospital Anxiety and Depression Scale (HADS)

The HADS is a 14-item self-report questionnaire comprising two 7-item subscales measuring anxiety and depression. It was originally developed as a clinical screening tool for use in a general medical outpatient setting and so explicitly excludes items that might be confounded by somatic aspects of illness or serious mental disorders.¹

Montreal Cognitive Assessment (MoCA)

The MoCA is a brief 30-item cognitive assessment tool designed to be sensitive to mild cognitive impairment in individuals presenting with subjective cognitive complaints.² The MoCA has previously been applied to chronic obstructive pulmonary disease (COPD) cohorts where it has been shown to be sufficiently sensitive to detect mild cognitive impairment in COPD patients with moderate-severe disease.³

Image acquisition

All images were acquired with a 3-Tesla Siemens Magnetron Skyra MRI scanner equipped with a 32-channel head coil with a maximum gradient strength of 45 mT/m. Sagittal T1-weighted 3D volume (T1W) images were acquired using a magnetization prepared rapid gradient echo sequence (TE=2.25 ms, TR=1800 ms, TI=800 ms, flip angle 9°, 169 contiguous sagittal slices with a 0.9 mm³ isotropic voxel dimension and field-of-view of 225 mm×240 mm×180 mm). Axial fluid-attenuated inversion recovery (FLAIR) was acquired using an inversion recovery sequence (TE-126 ms, TR=11,000 ms, TI=2690 ms, flip angle=150°, with 60 contiguous slices, voxel dimension of 0.7 mm×0.7 mm×3 mm and field-of-view of 201.25 mm×230 mm×180 mm). Diffusion tensor images (DTI) were acquired using an echo-planar imaging sequence with opposite phase-encode polarities (TE=76 ms, TR=6000 ms, flip angle =90°, 55 contiguous axial slices with a voxel dimension of 2.0 mm×2.0 mm×2.5 mm and field-of-view of 192 mm×192 mm×137.5 mm). For each phase-encode polarity, 8 volumes were acquired without diffusion sensitization and 60 with non-collinear diffusion gradients applied.

Image processing

Tissue macrostructure

The T1W images were re-sampled to 1 mm³ isotropic. The FLAIR was affine-registered to the T1W images using Advanced Normalisation Tools⁴ and a semi-automatic procedure used to segment the T1W images into supra-tentorial gray matter, white matter and cerebrospinal fluid tissue probability maps. White matter hyperintensities (WMHs) were segmented using the combined image intensities from the T1W and FLAIR images, then binarised at a manually determined threshold (i.e., dichotomized so that 1=WMH and 0=non-WMH). This process is described in full in Spilling et al, 2017 and Lambert et al, 2015.^5,6 Tissue volumes were quantified by integrating the values within each tissue segmentation and normalizing for head size – calculated as a percentage of total intracranial volume (gray matter + white matter + cerebrospinal fluid). Additionally, these tissue segmentations were used to define regions of normal-appearing white matter (NAWM) on the DTI (see below).

Tissue microstructure

The DTI data were corrected for movement artifacts, eddy-current distortions, and susceptibility-induced local gradients using FSL’s (FMRIB Software Library, version 5.0.6) “eddy”.⁷ The diffusion tensor model was fitted at every voxel within the DTI using FSL’s (FMRIB Software Library, version 5.0.6) “dtifit”,⁸ the skull removed using FSL’s (FMRIB Software Library, version 5.0.6) Brain Extraction Toolbox⁹ and mean diffusivity (MD) and fractional anisotropy (FA) maps calculated from the DTI, indicating the local magnitude and directionality of diffusion, respectively.

T1W images were aligned to the DTI data using the boundary-based registration procedure implemented in FSL’s (FMRIB Software Library, version 5.0.6) “epi-reg” script.¹⁰ This transformation was applied to the T1W tissue segmentations and binary WMH map (using trilinear interpolation) to align them with the DTI. The WMH map was re-binarised at 0.5 creating a WMH mask. These segmentations were used to define the probability of the DTI voxels belonging to each tissue-type. Voxels were considered to belong to the supra-tentorial NAWM where the probability of belonging to the white matter was higher than for any other tissue-type providing that they were not included within the WMH mask.

Normalized histograms (i.e. probability density functions) of FA and MD values within the NAWM were constructed using 100 equal bins ranging in value from 0 to 1 for FA and 0 to 2×10⁻⁴ mm²/s for MD. The median and peak height values were used to characterize the distribution of these histograms.

Acknowledgment

This study was funded by the British Lung Foundation. The funders had no role in the study design, data collection, analysis, interpretation or writing of the report.

Abbreviations

ANCOVA(s), analysis of covariance; BET, brain extraction tool; BMI, body mass index; CAT, COPD Assessment Test; COPD, chronic obstructive pulmonary disease; CRIC, Clinical Research Imaging Centre; CSF, cerebrospinal fluid; DTI, diffusion tensor imaging; FA, fractional anisotropy; FEV₁, forced expiratory volume in 1 Second; FLAIR, fluid attenuated inversion recovery; FVC, forced vital capacity; HADS, Hospital Anxiety and Depression Scale; IQR, interquartile range; MD, mean diffusivity; MoCA, Montreal Cognitive Assessment Test; MRI, magnetic resonance imaging; NAWM, normal-appearing white matter; PO_2, partial pressure of oxygen; PCO₂, partial pressure of carbon dioxide; SD, standard deviation; SaO_2, oxygen saturation; SVD, small vessel disease; TE, echo time; TI, inversion time; TIV, total intracranial volume; TR, repetition time; T1W, T₁-weighted; WMHs, white matter hyperintensities.

Author contributions

JWD and PWJ designed the study, JWD recruited the participants and acquired the clinical data. JWD and NJT acquired the MRI. CAS performed the analysis and drafted the manuscript. All authors contributed to data analysis, drafting and revising the article, gave final approval of the version to be published, and agree to be accountable for all aspects of the work.

Disclosure

PWJ is employed as a Global Medical Expert for GlaxoSmithKline. JWD reports grants from the British Lung Foundation, during the conduct of the study and has received personal fees and travel support unrelated to the content of this manuscript from Chiesi, Boehringer Ingelheim & NAPP pharmaceutical. DRB reports grants from National Institute for Health Research, during the conduct of the study. The authors report no other conflicts of interest in this work.