Acta Psychiatr Scand 2015: 131: 139–147 All rights reserved DOI: 10.1111/acps.12352

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd ACTA PSYCHIATRICA SCANDINAVICA

Neurobiological correlates of depressive symptoms in people with subjective and mild cognitive impairment
Auning E, Selnes P, Grambaite R, Saltyt_ e Benth J, Haram A, Løvli Stav A, Bjørnerud A, Hessen E, Hol PK, Muftuler løndalen A, Fladby T, Aarsland D. Neurobiological correlates of depressive symptoms in people with subjective and mild cognitive impairment. Objective: To test the hypothesis that depressive symptoms correlate with Alzheimer’s disease (AD) type changes in CSF and structural and functional imaging including hippocampus volume, cortical thickness, white matter lesions, Diffusion tensor imaging (DTI), and fluoro-deoxyglucose positron emission tomography (FDG-PET) in patient with subjective (SCI) and mild (MCI) cognitive impairment. Method: In 60 patients, depressive symptoms were assessed using the Geriatric Depression Scale. The subjects underwent MRI, 18F-FDG PET imaging, and lumbar CSF extraction. Results: Subjects with depressive symptoms (n = 24) did not have more pathological AD biomarkers than non-depressed. Uncorrected there were trends towards larger hippocampal volumes (P = 0.06), less orbital WM damage measured by DTI (P = 0.10), and higher orbital glucose metabolism (P = 0.02) in the depressed group. The findings were similar when SCI and MCI were analyzed separately. Similarly, in patients with pathological CSF biomarkers (i.e., predementia AD, n = 24), we found that correlations between scores on GDS and CSF Aß42 and P-tau indicated less severe AD-specific CSF changes with increasing depression. Conclusion: Depressive symptoms are common in SCI/MCI, but are not associated with pathological imaging or CSF biomarkers of AD. Depression can explain cognitive impairment in SCI/MCI or add to cognitive impairment leading to an earlier clinical investigation in predementia AD.

E. Auning1,2, P. Selnes3,


R. Grambaite3,4, J. Saltyt e_ 2,5 2,6 Benth , A. Haram , A. Løvli Stav3, A. Bjørnerud7, E. Hessen3,8, P. K. Hol7, A. Muftuler løndalen9, T. Fladby3,2, D. Aarsland1,10,11 1

Department of Geriatric Psychiatry, Akershus University Hospital, Ahus campus, Lørenskog, 2Institute of Clinical Medicine, Ahus campus University of Oslo, Oslo, 3 Department of Neurology, Akershus University Hospital, Lørenskog, 4Department of Child and Adolescent Mental Health Services, Akershus University Hospital, Grorud Outpatient Clinic, Oslo, 5HØKH, Research Centre, Akershus University Hospital, Oslo, 6Department of Geriatric Psychiatry, Østfold Central Hospital, Fredrikstad, 7The Intervention Centre, Oslo University Hospital, Rikshospitalet, Oslo, 8Department of Psychology, University of Oslo, Oslo, 9Department of Radiology and Nuclear Medicine, Oslo University Hospital, Radiumhospitalet, Oslo, Norway, 10Department of Neurobiology, Care Sciences and Society, Alzheimer’s Disease Research Centre, Karolinska Institutet, Stockholm, Sweden and 11Centre for Age-Related Diseases, Stavanger University Hospital, Stavanger, Norway Key words: subjective cognitive impairment; mild cognitive impairment; depression; Alzheimer’s disease; imaging Eirik Auning, Department of Geriatric Psychiatry, Akershus University Hospital, 1478 Lørenskog, Norway. E-mail: [email protected]

Accepted for publication October 2, 2014

Significant outcomes

• Depressive • •

symptoms are common in patients seeking help at memory clinics and can explain cognitive impairment. Patients with predementia Alzheimer’s disease may seek help at an early stage because of their depressive symptoms. The findings do not support the hypothesis that depression is associated with more severe Alzheimertype pathology in elderly people with subjective or mild cognitive impairment.

139

Auning et al.

Limitations

• A relatively small sample size. • Cross-sectional design. • Subjects with major depression were excluded. It is possible that we would have found biological correlates to depression if included patients were more depressed.

Introduction

Depression and dementia are common syndromes with considerable impact on patients and caregivers. Lifetime prevalence of depression has recently been estimated to about 13% worldwide and 19% in the United States (1). At present more than 30 million people worldwide are estimated to suffer from Alzheimers’s disease (AD), the most common disorder causing dementia, a number expected to raise considerably the next decades (2). Among known risk factors for dementia are mild cognitive impairment (MCI) (3) and depression (1). Early detection of AD using biomarkers reflecting disease progression is important. At present, there is solid evidence for biomarkers including traditional magnetic resonance imaging (MRI), fluoro-deoxy-glucose-positron emission tomography (FDG-PET), amyloid PET and cerebrospinal fluid (CSF) (4), and more recently also for indices derived from diffusion tensor imaging (DTI) (5, 6). Based on these findings, criteria for MCI due to AD, including biomarkers, have been proposed (4). In addition, the term subjective cognitive impairment (SCI) has received attention as a potential prodromal state of AD occurring before MCI (7, 8). Depression has been shown to be a common feature in SCI, MCI, and AD dementia (9–11). In a recent review (12), frequencies of comorbid MCI and depression in hospital based populations varied considerably, but the median frequency was shown to be as high as 44%. The mechanisms underlying the association between depression and AD are not known. There is some evidence from autopsy-studies that AD-type pathology are more pronounced in patients with comorbid depression or a lifetime history of major depression (13, 14), but other studies have not supported this hypothesis (9). Even less is known about the mechanisms underlying depression in predementia AD. Several studies have explored how biomarkers typically associated with AD are related to depression in non-demented and demented middle-aged and old people. Findings from amyloid imaging studies have been inconclusive, and 140

CSF studies have not found any association between AD pathology and depression (9), not even in depressed patients with primary dementia (15). In contrast, reduced hippocampal volume has been consistently observed in major depressive disorders (16) and functional neuroimaging studies, including FDG-PET, have reported that frontal hypometabolism is associated with depression in AD (17, 18) and MCI (19). However, at what stage hippocampal atrophy and frontal affection occur in the course of depression and their association with depression and cognitive impairment in AD is uncertain. In addition, DTI studies have shown lower anisotropy (arguably construed as more white matter damage) in frontal and temporal white matter tracts to be associated with depression (20). This supports the hypothesis that distinctive regional networks play a central role in depressive symptomatology and that frontotemporal subcortical neuronal projections are essential. Taken together, depressive symptoms are common and early features of AD often associated with frontal dysfunction, but disease mechanisms are unclear. Depression might be related to the brain changes associated with AD, but it may also reflect a secondary psychological reaction to the functional impairment associated with neurodegenerative disease, result from comorbid chronic ischemic disease or independently imitate cognitive impairment (pseudodementia). Further, depression has been described at AD stages preceding dementia, although criteria encompassing biomarkers and imaging excluding coexisting ischemic disease were not always employed (21). We recently found that cognitive complaints in patients with SCI and MCI were associated with minor depressive symptoms, but not CSF biomarkers and that depressive symptoms only to a limited extent were associated with objective cognitive impairment (10). To investigate these relationships further, we therefore included patients with SCI and MCI from a hospital based memory clinic. Patients completed lumbar puncture for CSF analysis, MRI including DTI, FDG-PET, and the Geriatric

Neurobiological correlates of depressive symptoms Depression Scale (22). Participants are part of the same cohort and largely overlapping with included patients in a recent study (10). Aims of the study

Based on the assumption that depression and AD share common etiologies and reports of involved frontosubcortical neuronal circuits in depressed patients, we hypothesized that (minor) depressive symptoms correlate with AD type changes in cerebrospinal fluid, structural and functional imaging including hippocampal volumes, cortical thickness, diffusion tensor imaging, and fluoro-deoxy-glucose positron emission tomography.

Material and methods Selection of subjects

Patients with SCI or MCI were recruited consecutively from referrals to a university-hospital-based memory clinic between 2006 and 2013 (6, 10). Inclusion criteria for both groups were age 40–79 and impaired cognition (SCI or MCI) for at least 6 months. Exclusion criteria were impaired activities of daily living (i.e., dementia), a previous diagnosis of a major psychiatric disorder (psychosis), cancer, drug abuse, solvent exposure, anoxic brain damage, including clinical stroke and/or cerebral infarction on imaging, or other severe physical disease which may influence cognition. Patients fulfilling ICD-10 criteria for a major depressive episode after a clinical interview in accordance with ICD-10 criteria and guidelines were excluded, as this is inconsistent with a diagnosis of MCI. Patients who fulfilled any of the criteria for dementia with Lewy bodies or frontotemporal dementia (23, 24) were also excluded. Clinical assessment

SCI and MCI were defined as the second (SCI) and third (MCI) stages of the Global Deterioration Scale (25, 26). Global Deterioration Scale staging was determined after a clinical interview and the following screening tests: Mini-Mental Status Examination (MMSE) (27), stepwise comparative status analysis (STEP) parameters 13–20 (28); I-Flex (fluency, interference, and numeral-letter items) (28) and Cognistat (29) (memory, including cued recall, and executive functions). In addition, the clinical dementia rating (CDR) (30) was administered by a trained rater based on an interview with the patient and a relative. This method

of determining cognitive stage is in agreement with a previous study (31). Depressive symptoms were assessed using the 15-item Geriatric Depression Scale (32) with 5/6 as cut off, a valid and reliable measure of depression in the elderly (33, 34) including populations with MCI (35). The Geriatric Depression Scale was applied as a screening method to assess potentially relevant clinical depressive symptoms. Thus, patients with Geriatric Depression Scale scores above cutoff (i.e., 6 points or more), not fulfilling ICD-10 criteria for major depression, were diagnosed with minor depression and included in the group with ‘depressive symptoms’ (n = 24). Patients with 5 points or less were included in the group without depressive symptoms (n = 36). A composite cerebrovascular risk score was made consisting of six items (smoking, diabetes, hypertension (i.e., blood pressure above 140/ 90 mmHg on two or more occasions and/or use of antihypertensiva), hypercholesterolemia (i.e., total cholesterol above 7 mmol/l or use of anticholesterol agents), hyperhomocysteinemia (i.e., above 15 lmol/l) and known cerebrovascular disease (other than stroke) and family history of cerebrovascular disease). The total cerebrovascular burden in the individual patient was thus assessed on a scale from 0 to 6, where 0 indicated no burden and 6 indicated maximum burden of disease. To adjust for potential confounding of cerebrovascular disease, we also included analysis of white matter (WM) lesions (WML). Biomarker assessment

CSF. After lumbar puncture, CSF AD biomarkers (Ab42, T-tau, P-tau) were analyzed according to protocol and as previously described (36). For CSF to be assessed as pathological, either Ab42 and/or P-tau had to be outside the specified reference area. MRI. MRI scans were obtained using a Siemens Espree 1.5 T system, one MPRAGE sequence was acquired (TR/TE/TI/FA = 2400/3.65/1000/8°, matrix = 240 9 192), 160 sagittal slices, thickness = 1.2 mm, in-plane resolution of 1 mm 9 1.2 mm. The protocol also included 2D axial fluid-attenuated inversion recovery (FLAIR) images with the following parameters: TR/TE/ TI = 13420/121/2500, 36 slices, spaced at 3.0, and 3.9 mm thick. The pulse sequences for DTI were b = 750; 12 directions repeated five times; 5 b0-values per slice, TR = 6100 ms, TE = 117 ms, number of slices: 30, slice thickness: 3 mm (gap 1.9 mm), in-plane

141

Auning et al. resolution: 1.2 9 1.2 mm2, bandwidth: 840 Hz/ pixel. The Oxford Centre for Functional MRI of the Brain (FMRIB) Software Library version 4.1 was used for DTI analyses and calculations. Initially, FMRIB’s Linear Image Registration Tool (37) was used for affine registrations of each DTI volume to the low-b (b = 0) image. Fractional anisotropy (FA) and eigenvalue maps were created. Radial diffusivity (DR) is defined as the mean of eigenvalue 2 and 3. Tract-based spatial statistics (38) was used for voxel-wise statistical analysis of the DTI variables (FA, DR). FMRIB’s Diffusion Toolbox was used to create DTI images by fitting a tensor model to the raw diffusion data, and Software Library version’s Brain Extraction Tool was used for subsequent brain extraction. All subjects’ FA data were then aligned into a common space using a non-linear registration tool (FMRIB), which uses a b-spline representation of the registration warp field (39). Further, the mean FA image was created and thinned to create a mean FA skeleton that represents the centers of all tracts common to the group. Each subject’s aligned FA data were then projected onto this skeleton and the resulting data fed into voxel-wise cross-subject statistics. DR data were then extracted from each subject according to the skeletonized FA map. Moreover, WM ROIs based on the FreeSurfer WM parcellations were extracted for FA and DR: The FSL FMRIB FA template (to which every subjects FA volume initially was registered) was co-registered to the standard space T1 volume MNI152, which subsequently went through the FreeSurfer processing stream to create a volume with WM parcellations. The processing stream includes segmentation of the subcortical WM and deep gray matter volumetric structures (40) and parcellation of the cortical surface (41) according to a previously published parcellation scheme (42). This labels cortical sulci and gyri, and thickness values are calculated in the ROIs. Based on the cortical parcellation, WM in the gyrus underneath each cortical label was identified. Each WM voxel within a gyrus was labeled according to the label of the nearest cortical voxel. Deep WM was not assigned to a particular cortical area, with a 5 mm distance limit. The registration between the FA template and the MNI152 volume was applied to the volume with the WM parcellations, and the resulting volume was used to extract the skeletonized DR data from each WM ROI. DR and FA were determined in WM ROIs using tract-based spatial statistics and whole brain DTI analysis (with threshold-free cluster enhancement to correct for multiple comparisons) was used to 142

compare participants with and without depressive symptoms.WML were assessed using the method published by Fazekas (43) by one reviewer with good test-retest reliability (Pearson correlation coefficient 0.8 and 0.9 for periventricular and subcortical WM lesions respectively). To reduce interindividual variation due to head size, relative hippocampal volume was computed (per mille of total intracranial volume). All other morphometric ROI measures were thickness, and correction for head size is not appropriate. Measurements were averaged between hemispheres. FDG-PET scanning and analyses. 18F-FDG PET/ CT imaging was performed with a Siemens Biograph 16 PET/CT scanner (Siemens Healthcare). Subjects fasted for at least 4 h prior to imaging (water only), and plasma glucose had to be ≤8 mmol/l for FDG to be injected. After 10 min rest with eyes closed, subjects had an intravenous bolus of 200 +/ 10 MBq 18F-FDG injected and rested for 45–60 min before scanning. The PET acquisition was performed in 3D mode with one single axial position, duration 15 min. Attenuation and scatter corrections were performed. The images were reconstructed by an iterative technique (five iterations, eight subsets), using a Gaussian smoothing filter with full width at half maximum of 3.5 mm, zoom 1. The image format was 256 9 256. For each subject, FDG-PET frames were registered to the corresponding intensity-normalized MRI volume. PET activity was averaged within each ROI defined on the MRI and normalized to activity within the brainstem. Age, gender and cortical thickness were regressed out of all variables and the standardized residuals were used in further analyses. Measurements were averaged between hemispheres. Selection of cortical and white matter regions of interest

For the MRI and FDG-PET analyzes, seven regions of interests (ROIs) were selected a priori based on presumed anatomical predilection sites for depression as discussed above. Thus, we focused on the entorhinal, parahippocampal, lateral orbitofrontal, medial orbitofrontal, superior frontal, orbital and anterior cingulate cortical ROIs respectively (42). The localizations of the preplanned ROIs are illustrated in Fig. 1. Statistics

Demographic and clinical characteristics of patients were described as means and standard deviations (SD) or frequencies and percentages as

Neurobiological correlates of depressive symptoms

Fig. 1. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral-based regions of interest (ROI), by Desikan et al. (42). Presented are cortical representations of the seven preplanned ROIs, here shown in one hemisphere. The left picture illustrates the lateral view of the hemisphere and the right shows the medial view of the hemisphere. Red = the orbitofrontal cortex (lateral division), Blue = the orbitofrontal cortex (medial division), Yellow = superiorfrontal cortex, Light green = parahippocampal cortex, Dark green = entorhinal cortex, Purple = orbital cortex, brown = anterior cingulate cortex.

appropriate. Kolmogorov–Smirnov test for normality was used to assess the distribution of depression scores and parameters of CSF, MRI, WM lesions, cerebrovascular risk factors and FDG-PET. For group comparison, independent samples t-test, Mann–Whitney U-test, Chi-square test, Pearson or Spearman correlations were used as appropriate (Table 1). For further analyses, Hippocampus volume, cortical thickness, DTI and FDG-PET variables were adjusted for age and sex by a linear regression model. PET variables were in addition adjusted for cortical thickness. Residuals of a linear regression model were used for further analyses. As most of these variables are likely to be highly correlated, a multivariate analysis of variance (MANOVA) was performed for simultaneous comparison with cognitive impairment (SCI or MCI) as factors in the model together with depressed/non-depressed (independent variables). The assumptions of MANOVA were assessed by performing Levene’s test for homogeneity of variance in each variable, multivariate box’s test for equality of covariance matri-

ces, and the Bartlett’s test for sphericity of correlation matrix. Wilk’s Lambda statistics was calculated to assess the difference in dependent variables between the depressed and non-depressed patients. Post hoc comparisons were performed by ANOVA. Analyzes stratified by CSF were defined as secondary and were performed by including CSF (pathological or not) in addition to cognitive classification (SCI or MCI) as factors into the MANOVA model together with depressed/non-depressed (independent variables) and interaction between the two. Bonferroni adjustment for multiple tests was applied for primary analyses only. Statistical analyses were performed by SPSS version 20 (IBM Corporation, Armonk, NY, USA). All tests were two-sided. P-values below 0.05 were considered significant. Ethical considerations

The study protocol was approved by the south eastern Norway ethical committee for medical research, and informed consent was obtained from

Table 1. Demographics and clinical characteristics Cognition

Depressive symptoms†

Variables

All subjects (n = 60)

SCI (n = 22)

MCI (n = 38)

Yes (n = 24)

No (n = 36)

Age, years Sex, female, n (%) MMSE score Geriatric depression scale score CSF (Pathologic Aß42 and/or p-tau) Cerebrovascular composite score Fazekas; mean white matter score Fazekas; mean periventricular score SCI/MCI, n

60.0  6.8 (45–76) 29 (48.3%) 28.2  1.4 (24–30) 5.4  3.7 (0–15) 24 (40.0%) 1.5  1.0 (0–3) 0.85  0.8 (0–3) 1.1  0.8 (0–3) 22/38

58.5  6.5 (45–68) 13 (59.1%) *29.2  0.9 (28–30) 5.5  3.8 (0–14) 6 (27.3%) 1.4  1.0 (0–3) 1.0  0.9 (0–3) 1.1  0.8 (0–3)

60.8  6.6 (48–76) 16 (42.1%) 27.7  1.5 (24–30) 5.4  3.7 (0–15) 18 (47.4%) 1.6  1.0 (0–3) 0.8  0.8 (0–2) 1.1  0.8 (0–3)

60.0  6.5 (48–76) 13 (50.0%) 28.1  1.2 (26–30) *9.2  2.7 (6–15) 7 (29.1%) 1.7  0.9 (0–3) 1.0  0.9 (0–3) 1.1  1.0 (0–3) 8/16

59.9  7.1 (45–69) 16 (47.1%) 28.3  1.5 (24–30) 2.9  1.6 (0–5) 17 (47.2%) 1.4  1.0 (0–3) 0.8  0.8 (0–2) 1.1  0.6 (0–2) 14/22

Numbers represent mean  standard deviation (range) if not indicated otherwise. Composite score from 0 to 6 was 0 which indicates no cerebrovascular burden, 6 indicates maximum disease burden. SCI, subjective cognitive impairment; MCI, mild cognitive impairment; MMSE, Mini-Mental State Examination. †Greater than five points on the Geriatric Depression Scale. *Significant at the P < 0.001 level.

143

Auning et al. all subjects before any study-specific procedures were performed. Results

This study included 60 participants (22 SCI and 38 MCI), and demographics and clinical characteristics are shown in Table 1. SCI and MCI patients did not differ except MCI were more cognitively impaired as expected. MCI had significantly more pathology on morphometry and DTI, but not CSF or PET findings (results not shown). Subjects with (n = 24) or without (n = 36) depressive symptoms did not differ for other clinical or demographic variables (Table 1). Biomarker changes in patients with and without depressive symptoms

Hippocampus volume, cortical thickness, DTI and FDG-PET. There were no regions with more pathological changes in the depressed compared to the non-depressed group. In contrast, there was a trend towards higher hippocampus volume (P = 0.06), lower orbital DR (less WM damage) (P = 0.10) and significantly higher orbital glucose metabolism in depressed patients compared to those without depressive symptoms (P = 0.02) before correction for multiple comparisons (results not shown). However, after Bonferroni adjustment for multiple tests, no significant differences were seen between the depressed and non-depressed group. These findings did not differ when SCI and MCI were assessed separately or when we studied only those with pathological CSF AD markers. Results from whole brain DTI analysis revealed no significant associations with depressive symptoms, but uncorrected there were multiple scattered areas in all brain regions showing less WM damage in patients with depressive symptoms (P = 0.05). There were no significant differences in cortical thickness in the two groups.

In patients with pathological CSF biomarkers (i.e., predementia AD, n = 24) we found correlation at trend-level between scores on the Geriatric Depression Scale and CSF Aß42 (r = 0.40, P = 0.05) (Fig. 2) and P-tau (r = 0.36, P = 0.08). Both correlations indicated less severe AD-specific CSF changes with increasing depression. The same tendency was seen when all subjects were analyzed together (but P > 0.10). Discussion

Our main aim was to find neurobiological correlates of mild depressive symptoms in people with SCI and MCI, focusing on AD-related biomarkers, to provide insights into common mechanisms of depression and AD. Our results suggest that AD-type brain changes are not associated with symptoms of depression in SCI and MCI. These findings did not differ when SCI and MCI were assessed separately or when we studied only those with pathological CSF AD markers, indicating that in predementia AD, mild depressive symptoms are not associated with pathological AD changes. In contrast, we found a tendency to less pronounced changes in those with depression. Taken together, our findings indicate that depressive symptoms are driving the cognitive symptoms. This is consistent with our previous findings that cognitive complaints in patients with SCI and MCI are associated with minor depressive symptoms (10). Similarly, Kramberger et al. (15) recently showed that CSF AD biomarkers are not related to depression in SCI and MCI. Longitudinal studies in people without AD have found that the severity of baseline WMLs predict

CSF

We found numerically more non-depressed patients compared to depressed patients with pathological CSF biomarkers, but the difference was not statistically significant (P = 0.19). There were no statistically significant differences between depressed and non-depressed patients on the dependent variables Aß42 and P-tau (Wilks Lambda = 0.67). 144

Fig. 2. Correlation between scores on the Geriatric Depression Scale and CSF Aß42 in predementia AD.

Neurobiological correlates of depressive symptoms depression at follow-up examination (44). This is in accord with the vascular depression hypothesis that states an association between depression and WMLs as well as cerebovascular risk factors such as hypertension, diabetes, and smoking (45) in mainly frontal subcortical regions (46–49). The vascular depression hypothesis, however, is challenged because chronic non-vascular illness also seems to be associated with late-life depression (50). We found that Fazekas and cerebrovascular scores were low and did not differ between those with and without depressive symptoms, suggesting that minor depressive symptoms in SCI and MCI are not driven by cerebrovascular disease. Other studies suggest that biological factors such as enlargement of the ventricles, reduced frontal lobe and hippocampus volumes, damaged WM tracts, and reduced blood flow to the prefrontal cortex (51) are important in non-AD depression, but how these changes interact with the development of cognitive impairment is unclear. A potential common link between depression and AD is the effect of chronic stress on the immune system including outflow of cytokines, influence on the Hypothalamic-Pituitary-Adrenal (HPA)-axis and secondary neurodegenerative effects on the hippocampus including effects on neurotrophic factors (52). Relatively few studies have explored the brain correlates of depression in people with AD, and sample sizes are often low and findings inconsistent. Frontal hypometabolism has been shown in several studies (17, 18), and there is some evidence that dopaminergic and serotonergic systems, the APOE e4 genotype, brain-derived neurotrophic factor, genetic variations, activation of the HPAaxis and neuroinflammation are associated with depression (9, 52). In a recent study, Lebedeva et al. (53) found that depressive symptoms in AD are associated with cortical thinning in temporal and parietal regions. Studies of cerebrospinal fluid, however, have not found any association between AD and depression (9, 15). Of note, most studies have included people with dementia of mild or moderate severity, and very few studies have explored the brain correlates of depression in people with predementia AD. In the paper by Grambaite et al. (10) including 70 SCI/ MCI patients from the same cohort as the present study, depressive symptoms, but not CSF biomarkers, were significantly associated with cognitive complaints, which is in accord with previous studies. In addition, we found a tendency towards less pathological CSF biomarkers with increasing depression also when subjects with pathological CSF (predementia AD, n = 24) were analyzed separately. This suggests that subjects with early

stage AD and depressive symptoms seek help at the memory clinic earlier than those without depressive symptoms. Also, major depression is inconsistent with a diagnosis of SCI/MCI, and biological changes are to some extent believed to be proportional to the severity of depression. In minor depressive states psychosocial factors are more likely to cause and maintain depressive symptoms, with a subsequent higher tendency for symptoms to fluctuate, and the biological changes are thus probably less pronounced than in more advanced cases. This may in part explain why depressed patients in this SCI/ MCI cohort had less pathological AD biomarkers than non-depressed. There is also evidence that treatment with antidepressants can increase hippocampal cell proliferation and neurogenesis (54). In addition to our finding that the depressed have less neurodegenerative changes, past and present antidepressant use could be another explanation for the less hippocampal changes in depressed compared to nondepressed subjects. At baseline examination, only seven patients (five scored 6 points or more on the Geriatric Depression Scale) used antidepressants, but it is unlikely that this has significantly influenced the results. However, earlier use of antidepressants was not recorded. In contrast, a recent study found that antidepressant use in mild AD and dementia with Lewy bodies was associated with parahippocampal thinning (55) confirming that the neurobiology of antidepressants, including possible neuroprotective properties, is complex and at present unclear. There are other possible explanations to cognitive impairment in SCI and MCI, including cerebrovascular disease and unspecific activation of the immune system, and future research should focus on how to differentiate between the different causes to provide more targeted treatment. Limitations and strengths

• •

•

It is difficult to detect small effect sizes when sample size is low, exposing the method to type II error. Subjects with major depression were excluded as it is inconsistent with a diagnosis of MCI. However, it is possible that we would have found biological correlates to depression if included patients were more depressed. SCI and MCI are heterogeneous states and differences between (minor) depressed and nondepressed subjects may have been masked by other overlapping conditions such as other dementia subtypes. On the other hand, the 145

Auning et al. included patients fulfilled consensus criteria for SCI/MCI and not any of the core or suggestive criteria for dementia with Lewy bodies or criteria consistent with a diagnosis of frontotemporal dementia. • A cross-sectional design is not optimal as depressive symptoms, in particular in minor depression, may fluctuate and not be present at follow-up examination. A longitudinal approach will help to clarify the association between AD markers and risk for future depression. To conclude, the neural mechanisms underlying dementia and depressive disorders are complex, and clarification is important to improve future assessment, early intervention and treatment. Our findings suggest that in people with SCI and MCI, in particular in those with normal CSF or imaging, depression may explain the cognitive impairment. A thorough examination of depressive symptoms in memory clinic patients is therefore warranted. On the other hand, a subgroup of patients have both depressive symptoms and pathological AD biomarkers and may seek help at the memory clinic at an earlier stage than non-depressed. Our findings suggest that depression may add to cognitive impairment leading to an earlier clinical investigation in predementia AD. Declaration of interest
Eirik Auning, Per Selnes, Ramune Grambaite, J urat_e Saltyt_e Benth, Astrid Haram, Ane Løvli Stav, Atle Bjørnerud, Erik Hessen, Per Kristian Hol and Tormod Fladby have nothing to declare. Ayca Muftuler løndalen serves on the advisory board of Vintafolide/Etarfolide MSD Norway, a subsidiary of Merck & Co, U.S.A. Dag Aarsland has received honoraria or research support from Lundbeck, Inc., Novartis, GE Healthcare, and GlaxoSmithKline; serves on the editorial boards of International Psychogeriatrics, Movement Disorders, Gerontology, and the Journal of Neurology, Neurosurgery and Psychiatry.

References 1. Barnes DE, Yaffe K. The projected effect of risk factor reduction on Alzheimer’s disease prevalence. Lancet Neurol 2011;10:819–828. 2. Ferri CP, Prince M, Brayne C et al. Global prevalence of dementia: a Delphi consensus study. Lancet 2005;366:2112–2117. 3. Petersen RC, Roberts RO, Knopman DS et al. Mild cognitive impairment: ten years later. Arch Neurol 2009;66:1447–1455. 4. Sperling RA, Aisen PS, Beckett LA et al. Toward defining the preclinical stages of Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 2011;7: 280–292.

146

5. Bozzali M, Cherubini A. Diffusion tensor MRI to investigate dementias: a brief review. Magn Reson Imaging 2007;25:969–977. 6. Selnes P, Aarsland D, Bjornerud A et al. Diffusion tensor imaging surpasses cerebrospinal fluid as predictor of cognitive decline and medial temporal lobe atrophy in subjective cognitive impairment and mild cognitive impairment. J Alzheimers Dis 2013;33:723–736. 7. Stewart R, Godin O, Crivello F et al. Longitudinal neuroimaging correlates of subjective memory impairment: 4-year prospective community study. Br J Psychiatry 2011;198:199–205. 8. Scheef L, Spottke A, Daerr M et al. Glucose metabolism, gray matter structure, and memory decline in subjective memory impairment. Neurology 2012;79:1332–1339. 9. Enache D, Winblad B, Aarsland D. Depression in dementia: epidemiology, mechanisms, and treatment. Curr Opin Psychiatry 2011;24:461–472. 10. Grambaite R, Hessen E, Auning E et al. Correlates of subjective and mild cognitive impairment: depressive symptoms and CSF biomarkers. Dement Geriatr Cogn Dis Extra 2013;3:291–300. 11. Shahnawaz Z, Reppermund S, Brodaty H et al. Prevalence and characteristics of depression in mild cognitive impairment: the Sydney Memory and Ageing Study. Acta Psychiatr Scand 2013;127:394–402. 12. Panza F, Frisardi V, Capurso C et al. Late-life depression, mild cognitive impairment, and dementia: possible continuum? Am J Geriatr Psychiatry 2010;18:98–116. 13. Rapp MA, Schnaider-Beeri M, Purohit DP et al. Increased neurofibrillary tangles in patients with Alzheimer disease with comorbid depression. Am J Geriatr Psychiatry 2008;16:168–174. 14. Rapp MA, Schnaider-Beeri M, Grossman HT et al. Increased hippocampal plaques and tangles in patients with Alzheimer disease with a lifetime history of major depression. Arch Gen Psychiatry 2006;63:161–167. 15. Kramberger MG, Jelic V, Kareholt I et al. Cerebrospinal fluid Alzheimer markers in depressed elderly subjects with and without Alzheimer’s disease. Dement Geriatr Cogn Dis Extra 2012;2:48–56. 16. McKinnon MC, Yucel K, Nazarov A et al. A meta-analysis examining clinical predictors of hippocampal volume in patients with major depressive disorder. J Psychiatry Neurosci 2009;34:41–54. 17. Hirono N, Mori E, Ishii K et al. Frontal lobe hypometabolism and depression in Alzheimer’s disease. Neurology 1998;50:380–383. 18. Holthoff VA, Beuthien-Baumann B, Kalbe E et al. Regional cerebral metabolism in early Alzheimer’s disease with clinically significant apathy or depression. Biol Psychiatry 2005;57:412–421. 19. Lee HS, Choo IH, Lee DY et al. Frontal dysfunction underlies depression in mild cognitive impairment: a FDG-PET study. Psychiatry Investig 2010 Sep;7:208–214. 20. Sexton CE, Mackay CE, Ebmeier KP. A systematic review of diffusion tensor imaging studies in affective disorders. Biol Psychiatry 2009;66:814–823. 21. Amieva H, Le GM, Millet X et al. Prodromal Alzheimer’s disease: successive emergence of the clinical symptoms. Ann Neurol 2008;64:492–498. 22. Burke WJ, Roccaforte WH, Wengel SP. The short form of the Geriatric Depression Scale: a comparison with the 30-item form. J Geriatr Psychiatry Neurol 1991;4:173–178.

Neurobiological correlates of depressive symptoms 23. McKeith IG, Dickson DW, Lowe J et al. Diagnosis and management of dementia with Lewy bodies: third report of the DLB Consortium. Neurology 2005;65:1863–1872. 24. THE LUND AND MANCHESTERGROUPS. Clinical and neuropathological criteria for frontotemporal dementia. J Neurol Neurosurg Psychiatry 1994;57:416–418. 25. Auer S, Reisberg B. The GDS/FAST staging system. Int Psychogeriatr 1997;9(Suppl):167–71. 26. Reisberg B, Ferris SH, DE Leon MJ, Crook T. Global Deterioration Scale (GDS). Psychopharmacol Bull 1988;24: 661–663. 27. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12: 189–198. 28. Wallin A, Edman A, Blennow K et al. Stepwise comparative status analysis (STEP): a tool for identification of regional brain syndromes in dementia. [Erratum appears in J Geriatr Psychiatry Neurol 1997 Apr, 10(2), pp. 90–1]. J Geriatr Psychiatry Neurol 1996;9: 185–199. 29. Kiernan RJ, Mueller J, Langston JW et al. The neurobehavioral cognitive status examination: a brief but quantitative approach to cognitive assessment. Ann Intern Med 1987;107:481–485. 30. Morris JC. Clinical dementia rating: a reliable and valid diagnostic and staging measure for dementia of the Alzheimer type. Int Psychogeriatr 1997;9(Suppl):173–176. 31. Nordlund A, Rolstad S, Hellstrom P et al. The Goteborg MCI study: mild cognitive impairment is a heterogeneous condition. J Neurol Neurosurg Psychiatry 2005;76:1485– 1490. 32. Yesavage JA, Brink TL, Rose TL et al. Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res 1982;17:37–49. 33. Montorio I, Izal M. The Geriatric Depression Scale: a review of its development and utility [Review] [50 refs]. Int Psychogeriatr 1996;8:103–112. 34. Almeida OP, Almeida SA. Short versions of the geriatric depression scale: a study of their validity for the diagnosis of a major depressive episode according to the ICD-10 and DSM-IV. Int J Geriatr Psychiatry 1999;14: 858–865. 35. Debruyne H, Van BM, Le BN et al. Is the geriatric depression scale a reliable screening tool for depressive symptoms in elderly patients with cognitive impairment? Int J Geriatr Psychiatry 2009;24:556–562. 36. Sjogren M, Vanderstichele H, Agren H et al. Tau and Abeta42 in cerebrospinal fluid from healthy adults 21–93 years of age: establishment of reference values. Clin Chem 2001;47:1776–1781. 37. Woolrich MW, Jbabdi S, Patenaude B et al. Bayesian analysis of neuroimaging data in FSL. Neuroimage 2009;45(Suppl):173–186. 38. Smith SM, Jenkinson M, Johansen-Berg H et al. Tractbased spatial statistics: voxelwise analysis of multi-subject diffusion data. Neuroimage 2006;31:1487–1505. 39. Rueckert D, Sonoda LI, Hayes C et al. Nonrigid registration using free-form deformations: application to breast

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

MR images. IEEE Trans Med Imaging 1999;18: 712–721. Fischl B, Salat DH, Busa E et al. Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron 2002;33:341–355. Fischl B, van der Kouwe A, Destrieux C et al. Automatically parcellating the human cerebral cortex. Cereb Cortex 2004;14:11–22. Desikan RS, Segonne F, Fischl B et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage 2006;31:968–980. Fazekas F, Chawluk JB, Alavi A et al. MR signal abnormalities at 1.5 T in Alzheimer’s dementia and normal aging. AJR Am J Roentgenol 1987;149:351–356. Herrmann LL, Le MM, Ebmeier KP. White matter hyperintensities in late life depression: a systematic review. J Neurol Neurosurg Psychiatry 2008;79:619–624. Sneed JR, Culang-Reinlieb ME. The vascular depression hypothesis: an update. Am J Geriatr Psychiatry 2011;19:99–103. Greenwald BS, Kramer-Ginsberg E, Krishnan KR et al. Neuroanatomic localization of magnetic resonance imaging signal hyperintensities in geriatric depression. Stroke 1998;29:613–617. Macfall JR, Payne ME, Provenzale JE et al. Medial orbital frontal lesions in late-onset depression. Biol Psychiatry 2001;49:803–806. Videbech P, Ravnkilde B, Gammelgaard L et al. The Danish PET/depression project: performance on Stroop’s test linked to white matter lesions in the brain. Psychiatry Res 2004;130:117–130. O’brien JT, Firbank MJ, Krishnan MS et al. White matter hyperintensities rather than lacunar infarcts are associated with depressive symptoms in older people: the LADIS study. Am J Geriatr Psychiatry 2006;14:834–841. Valkanova V, Ebmeier KP. Vascular risk factors and depression in later life: a systematic review and metaanalysis. Biol Psychiatry 2013;73:406–413. Crocco EA, Castro K, Loewenstein DA. How late-life depression affects cognition: neural mechanisms. Curr Psychiatry Rep 2010;12:34–38. Casanova MF, Starkstein SE, Jellinger KA. Clinicopathological correlates of behavioral and psychological symptoms of dementia. Acta Neuropathol (Berl) 2011; 122:117–135. Lebedeva A, Westman E, Lebedev AV et al. Structural brian changes associated with depressive symptoms in the elderly with Alzheimers‘s disease. J Neurol Neurosurg Psychiatry 2014;85:930–935. Pilar-Cuellar F, Vidal R, Diaz A et al. Neural plasticity and proliferation in the generation of antidepressant effects: hippocampal implication. Neural Plast 2013;2013: 537265. Lebedev AV, Beyer MK, Fritze F et al. Cortical changes associated with depression and antidepressant use in Alzheimer and Lewy body dementia: an MRI surface-based morphometric study. Am J Geriatr Psychiatry 2014;22:4– 13.

147