Plasma Amyloid- β and Alzheimer’s Disease-Related Changes in Late-Life Depression

Abstract

To elucidate an involvement of amyloid dysmetabolism in the pathophysiology of depression, we investigated associations of plasma amyloid-β (Aβ) levels with Alzheimer’s disease-related changes in neuroimaging and cognitive dysfunction in patients with late-life depression. Higher plasma Aβ₄₀, but not Aβ₄₂ nor Aβ₄₀/Aβ₄₂ ratio, was associated with higher degree of parahippocampal atrophy and lower verbal fluency performance. Indeed, high plasma Aβ₄₀ predicted poor cognitive prognosis of depressed patients with mild cognitive impairment. As an anti-depressive treatment, electroconvulsive therapy (ECT) resulted in a marginally significant reduction of plasma Aβ₄₀ compared to pharmacotherapy alone, suggesting protective effects of ECT against amyloid dysmetabolism.

Keywords

Alzheimer’s disease amyloid-β electroconvulsive therapy late-life depression mild cognitive impairment neuroimaging parahippocampal atrophy plasma biomarker

INTRODUCTION

Accumulating evidence from meta-analyses of cohort studies suggests that a history of depression approximately doubles a risk of developing Alzheimer’s disease (AD) later in life [1, 2]. Individuals with mild cognitive impairment (MCI) show a strikingly high conversion rate to AD when they are accompanied by depression [3]. The concept that late-life depression (LLD), MCI, and AD may represent a clinical continuum is consistent with the hypothesis of “amyloid-associated depression” [4]. This hypothesis has been originally proposed based on the findings that patients with LLD have a lower concentration of plasma amyloid-β peptide 42 (Aβ₄₂) combined with a higher concentration of plasma Aβ₄₀ compared to control elderly individuals [4], which is further supported by a recent meta-analysis [5]. Indeed, an association between high plasma Aβ₄₀/Aβ₄₂ ratio and an increased risk of AD has been shown by large population studies [6, 7]. Recently, a brain amyloid imaging study has revealed that LLD with MCI are characterized by higher Aβ burden [8]. Taken together, altered amyloid metabolism seems to be involved in the pathophysiology of LLD, in which antidepressants often demonstrate only modest efficacy [9]. Modulation of amyloid dysmetabolism may have a large impact on improving outcome in treatment-resistant LLD and preventing subsequent conversion to AD.

The purpose of the present study is to elucidate an involvement of amyloid dysmetabolism in the pathophysiology of LLD by investigating associations of plasma Aβ levels with AD-related changes in neuroimaging and cognitive dysfunction. In the context of the LLD-MCI-AD continuum, we discuss possibilities of plasma Aβ as a prognostic or predictive biomarker that can be modified by anti-depressive therapies including electroconvulsive therapy (ECT).

MATERIALS AND METHODS

Patients and study protocol

The study was approved by the Ethical Committee of the University of Yamanashi, and all the patients gave their written informed consent. We recruited a consecutive series of 42 patients aged 50 years and over who were inpatient of University of Yamanashi Hospital between 2010 and 2013. They were diagnosed as major depressive disorder, or in a depressive episode of bipolar I/bipolar II disorder according to the DSM-IV-TR criteria. None of the participants met the DSM-IV-TR criteria for dementia. All the patients received mono- or combination drug therapy; 60% of the patients received selective serotonin reuptake inhibitors (SSRIs), serotonin-noradrenalin reuptake inhibitors (SNRIs), or noradrenergic and specific serotonergic antidepressants (NaSSAs), 48% tricyclic/tetracyclic antidepressants, and 36% mood-stabilizers or atypical antipsychotics. The doses of antidepressants were converted to an equivalent dose of imipramine [10]. Out of the 42 patients, 13 patients (31%) received ECT by clinical decision. The situations of the patients were consistent with the recommendations for use of ECT by the American Psychiatric Association Task Force on ECT [11]. Under anesthesia with thiamylal sodium and muscular relaxation with suxamethonium, bilateral brief pulse ECT was performed with a Thymatron SYSTEM IV device (Somatics LLC, Lake Bluff, IL, USA). Details of the ECT procedures were described previously [12].

Within 2 weeks after admission and within 2 weeks before discharge, biochemical analysis of peripheral blood, neuropsychological assessments, as well as neuroimaging examination were performed. For the patients received ECT, all these examinations were done before the first ECT and within 2–4 weeks after the last ECT.

Blood sampling and analysis

Fasting morning blood samples were collected and centrifuged and plasma were stored at –80°C. A sandwich enzyme-linked immunosorbent assay (ELISA) kit (Wako Pure Chemical Industries, Osaka, Japan) was used to determine Aβ₄₀ and Aβ₄₂ plasma concentrations. Details of the ELISA kits with the specific monoclonal antibodies for Aβ₄₀ and Aβ₄₂ were described previously [13]. Apolipoprotein E (APOE) phenotype was determined directly from plasma by isoelectric focusing followed by immunoblotting [14].

Neuropsychological assessments

We administered the Geriatric Depression Scale (GDS), the Beck Depression Inventory-II (BDI-II), and the Hamilton Depression Rating Scale (HAM-D), as scales of depressive symptoms. As for scales of cognitive functions, we administered the Mini-Mental State Examination (MMSE), the Clinical Dementia Rating (CDR), the Logical Memory I and II subscales of the Wechsler Memory Scale-Revised (WMS-R), the Wisconsin Card Sorting Test, semantic verbal fluency test (VFT) for 3 categories and phonemic VFT for 3 letters. The patients were determined to be cognitively normal or at the stage of MCI by reference to the Alzheimer’s Disease Neuroimaging Initiative (ADNI) criteria defined with scores of the MMSE, the CDR, and the Logical Memory II subscale of the WMS-R, and years of education [15].

Neuroimaging examination and analysis

Neuroimaging examinations were performed by investigating brain magnetic resonance imaging (MRI) and brain perfusion single-photon emission tomography (SPECT). Statistical image analyses were performed with the voxel-based specific regional analysis system for AD (VSRAD) for T1-weighed image of 1.5 Tesla MRI [16] and the easy Z-score imaging system (eZIS) for technetium-99 m ethyl cysteinate dimmer (Tc99m-ECD) SPECT [17]. Indeed, Z-scores of the VSRAD system were determined to assess the degree of atrophy of the parahippocampal gyrus [16]. The eZIS system provided three indicators, eZIS-severity, -extent, and -ratio, assessing the degree of reduction of regional cerebral blood flow (rCBF) in the early-stage AD-specific regions including the posterior cingulate gyrus, precuneus, and parietal cortices [17].

Statistical analysis

Table 1

Demographic and clinical characteristics of the LLD patients according to treatment group

Parameters	Pharmacotherapy	Pharmacotherapy plus	Total
	alone group	ECT group	(n = 42)
	(n = 29)	(n = 13)
Age (years)^a	67.8±9.9	70.3±10.6	68.6±10.1
Female [n (%)]^b	22 (76)	9 (69)	31 (74)
Possession of
APOE4^b	5 (17)	1 (8)	6 (14)
Diagnosis [n (%)]^b
MDD	22 (76)	10 (77)	32 (76)
BP-I/BP-II	7 (24)	3 (23)	10 (24)
Antidepressants doses (mg)^a,c
on admission	112.2±91.8	105.3±79.5	110.1±87.3
prior to discharge	126.2±100.6	117.0±96.6	123.4±98.3
HAM-D^a
on admission	21.9±9.2	23.3±8.3	22.3±8.8
prior to discharge	10.4±7.0^**	7.7±4.4^**	9.4±6.2^**
MMSE^a
on admission	23.9±3.1^*	26.6±3.3^*	24.8±3.4
prior to discharge	27.6±2.9^**	29.2±1.4	28.0±2.7^**

APOE4, apolipoprotein E ɛ4 allele; BP-I, bipolar I disorder; BP-II, bipolar II disorder; ECT, electroconvulsive therapy; HAM-D, Hamilton Depression Rating Scale; LLD, late-life depression; MDD, major depressive disorder; MMSE, Mini-Mental State Examination. ^aData are shown as the mean±S.D. None of the differences between the two groups were significant, except for the MMSE scores on admission (^*p < 0.05) by the Mann-Whitney U test. ^bNone of the differences between the two groups were significant by χ² test. The HAM-D scores were significantly decreased between admission and discharge in the pharmacotherapy alone group, the pharmacotherapy plus ECT group, and all the patients by two-tailed paired t test (^**p < 0.0001). Also, the MMSE scores were significantly increased between admission and discharge in the pharmacotherapy alone group and all the patients. ^cDoses of antidepressants are converted to an equivalent dose of imipramine.

Statistics were performed using JMP v12 for Windows. Differences in continuous variables between two treatment subgroups were analyzed by nonparametric statistics using the Mann-Whitney U test, and those among three subgroups divided by cognitive profile were analyzed by analysis of variance (ANOVA) with post hoc Tukey’s multiple comparison test. The comparison of the parameters on admission with those prior to discharge was performed by two-tailed paired t test. Pearson’s correlation was used to evaluate the correlation between the parameters. Categorical data were analyzed using χ² test.

RESULTS

Demographic and clinical characteristics of the LLD patients are shown in Table 1. Improvement in depressive symptoms was shown by a significant decrease in the HAM-D scores between admission and discharge, which was observed in both of the treatment subgroups and all the patients. Also,the MMSE scores were significantly increased in the pharmacotherapy alone group and all the patients. The increase in MMSE observed in the pharmacotherapy plus ECT group was not significant, which might be due to a ceiling effect because of higher MMSE scores at baseline in this group.

The average plasma concentration of Aβ₄₀ and Aβ₄₂ of all the LLD patients on admission were 61.8±17.0 (S.D.) and 9.7±4.4 pmol/L, and those prior to discharge were 64.4±19.6 and 10.3±5.1 pmol/L, respectively. Both plasma Aβ₄₀ and Aβ₄₂ levels showed no significant changes between admission and discharge. No significant effects of possession of APOE4 allele were observed on plasma concentration of Aβ₄₀ and Aβ₄₂.

There were several significant associations between depression scales and AD-related parameters measured prior to discharge. Significant positive correlations were found between the HAM-D score or the GDS score and the VSRAD Z-score (an index of parahippocampal atrophy) (r = 0.49 and 0.48, respectively), and between the BDI-II scores and eZIS-severity (an index of the rCBF reduction in the AD-specific regions) (r = 0.74). Also, the BDI-II scores were correlated with plasma Aβ₄₀ levels measured prior to discharge (r = 0.52) (all p < 0.05, data not shown).

Levels of plasma Aβ₄₀ measured both on admission (r = 0.52, Fig. 1a) and prior to discharge (r = 0.50, data not shown) were positively correlated with the VSRAD Z-score. Moreover, plasma Aβ₄₀ levels measured prior to discharge were negativelycorrelated with phonemic VFT scores (r = –0.47, Fig. 1b) (all p < 0.05). None of the depression scales and parameters of cognitive function and neuroimaging were associated with Aβ₄₂ levels nor Aβ₄₀/Aβ₄₂ ratio in plasma.

Fig.1

Plasma Aβ₄₀ levels are associated with brain structural and cognitive alterations in the patients with late-life depression. a) Higher levels of plasma Aβ₄₀ examined on admission are associated with higher degree of atrophy of the parahippocampal gyrus measured as Z-scores of the MRI image-analysis (r = 0.52, p < 0.05). b) Even after the hospitalization therapy, higher levels of plasma Aβ₄₀ examined prior to discharge are associated with lower scores of phonemic verbal fluency performance (r = –0.47, p < 0.05).

According to their cognitive status at baseline and post-treatment, the LLD patients were divided into three groups using the ADNI criteria, i.e., subjects with stably normal cognition, MCI subjects reverted to normal with the anti-depressive treatment (MCI-reverters), and MCI subjects remained stable even after the treatment (MCI-non-reverters). Plasma Aβ₄₀ levels examined on admission were significantly higher in the MCI-non-reverters compared with the cognitively normal subjects or the MCI-reverters (Fig. 2, p < 0.01).

Fig.2

Levels of plasma Aβ₄₀ may predict cognitive outcome in patients with late-life depression. According to their cognitive status, the patients were divided into three groups; namely, cognitively normal subjects, mild cognitive impairment (MCI) subjects reverted to normal with the anti-depressive treatment (MCI-Reverters), and MCI subjects remained stable even after the treatment (MCI-Non-Reverters). ANOVA with post hoc Tukey’s multiple comparison test revels that plasma Aβ₄₀ levels examined on admission are significantly higher in the MCI-Non-Reverters compared with the cognitively normal subjects or the MCI-Reverters (^*p < 0.01). Data shown as mean with S.D.

When the patients treated with pharmacotherapy plus ECT were compared to those treated with pharmacotherapy alone, there were no significant differences between the two groups in baseline levels of plasma Aβ₄₀ (64.7±17.2 and 60.5±17.1 pmol/L, respectively) and Aβ₄₂ (10.6±3.4 and 9.3±4.7 pmol/L, respectively). However, a marginally significant reduction in plasma Aβ₄₀ was observed after anti-depressive therapy with the pharmacotherapy plus ECT compared to pharmacotherapy alone (Supplementary Figure 1, p = 0.059). Indeed, the numbers of ECTs (4–12 times) were associated with the improvement in scores of the HAM-D (r = 0.72), the BDI-II (r = 0.92) and phonemic VFT (r = 0.84). Also, the numbers of ECTs were negatively associated with the eZIS-severity (r = –0.76) and the eZIS-extent (r = –0.81) measured prior to discharge (all p < 0.05, data not shown).

DISCUSSION

We have observed that there is a link between severity of residual symptoms in the LLD patients after the anti-depressive treatment and brain structural/functional alterations that are associated with early-stage AD, which supports the hypothesis that LLD and AD represent a possible clinical continuum [4, 8].

In view of predicting conversion from LLD to AD, plasma Aβ₄₀ may have a significant role as a predictive biomarker. Indeed, a higher level of plasma Aβ₄₀ at baseline is associated with parahippocampal atrophy and poor cognitive prognosis. Furthermore, a higher level of plasma Aβ₄₀ after the anti-depressive treatment is associated with severity of residual symptoms in depression, cognitive dysfunction, and parahippocampal atrophy, indicating a possible link between amyloid dysmetabolism and treatment resistance in LLD. These associations of plasma Aβ with several parameters have been demonstrated only in Aβ₄₀, but not Aβ₄₂ nor Aβ₄₀/Aβ₄₂ ratio. This is not surprising, because this is consistent with findings of an 11-year follow-up study of an elderly cohort from the Rotterdam Study [18], where a longitudinal association between depression and increased plasma Aβ₄₀, but not Aβ₄₂, has been demonstrated in the participants who develop dementia during the follow-up period. Therefore, plasma Aβ₄₀ may have a potency reflecting Aβ dysmetabolism especially in the LLD patients at risk of AD. Because a brain MRI study of patients with MCI and AD has reported an association of plasma Aβ₄₀ levels with the extent of white matter hyperintensity and lacunar infarcts [19], plasma Aβ₄₀ may be a biomarker of microvascular damage that promotes LLD and worsens its outcomes [9]. Elucidation of the biological mechanisms underlying the connection between plasma Aβ₄₀ and LLD needs further investigations.

Another remarkable finding of the present study is possible protective effects of ECT on AD-related changes in the LLD patients, which may give an intriguing view of preventive strategy against conversion from LLD to AD. Indeed, ECT is recommended for use by the American Psychiatric Association Task Force on ECT for the severe depression [11, 20], and the safety of ECT for LLD has been reported regardless of a diagnosis of dementia, MCI, or no cognitive impairment [21].

It has been reported that the plasma Aβ₄₀ and Aβ₄₂ show a transient increase immediately after the ECT, followed by the normalization 2 hours after the ECT [22], and show no significant change 1 week after the ECT [23]. We measured plasma Aβ within 2–4 weeks after the ECT and found a tendency toward reduction of plasma Aβ₄₀ in the patients received ECT compared to those received pharmacotherapy alone. The number of stimulations was associated with an improvement of cognitive performance and an alleviation of the reduction of brain perfusion in the AD-specific regions, suggesting possible protective effects of ECT against AD-associated deterioration. Of note, the main limitations of this study include the small sample size and the observational design without random allocation to the treatments. The possible efficacy of ECT on AD-related parameters needs further investigations with a larger sample size and an interventional design.

In conclusion, high levels of plasma Aβ₄₀ might be an efficient peripheral biomarker reflecting abnormal Aβ metabolism that mediates the pathophysiological link between LLD and AD, which can be modified by brain stimulation.

Footnotes

ACKNOWLEDGMENTS

Work in the authors’ laboratories is supported by funding from the Japan Society for the Promotion of Science (Grant-in-Aid for Scientific Research (C) 26461741 to AN).

Authors’ disclosures available online ().

Appendix

References

Ownby

, Crocco

, Acevedo

, John

, Loewenstein

(2006) Depression and risk for Alzheimer disease: Systematic review, meta-analysis, and metaregression analysis. Arch Gen Psychiatry 63, 530–538.

Diniz

, Butters

, Albert

, Dew

, Reynolds

3rd (2013) Late-life depression and risk of vascular dementia and Alzheimer’s disease: Systematic review and meta-analysis of community-based cohort studies. Br J Psychiatry 202, 329–335.

Modrego

, Ferrández

(2004) Depression in patients with mild cognitive impairment increases the risk of developing dementia of Alzheimer type: A prospective cohort study. Arch Neurol 61, 1290–1293.

Sun

, Steffens

, Au

, Folstein

, Summergrad

, Yee

, Rosenberg

, Mwamburi

, Qiu

(2008) Amyloid-associated depression: A prodromal depression of Alzheimer disease? Arch Gen Psychiatry 65, 542–550.

Nascimento

, Silva

, Malloy-Diniz

, Butters

, Diniz

(2015) Plasma and cerebrospinal fluid amyloid-β levels in late-life depression: A systematic review and meta-analysis. J Psychiatr Res 69, 35–41.

van Oijen

, Hofman

, Soares

, Koudstaal

, Breteler

(2006) Plasma Aβ1-40 and Aβ1-42 and the risk of dementia: A prospective case-cohort study. Lancet Neurol 5, 655–660.

Graff-Radford

, Crook

, Lucas

, Boeve

, Knopman

, Ivnik

, Smith

, Younkin

, Petersen

, Younkin

(2007) Association of low plasma Aβ42/Aβ40 ratios with increased imminent risk for mild cognitive impairment and Alzheimer disease. Arch Neurol 64, 354–362.

, Liu

, Chen

, Hsiao

, Hsieh

, Lee

, Yen

, Lin

(2016) Beta-amyloid deposition and cognitive function in patients with major depressive disorder with different subtypes of mild cognitive impairment: (18)F-florbetapir (AV-45/Amyvid) PET study. Eur J Nucl Med Mol Imaging 43, 1067–1076.

Mahgoub

, Alexopoulos

(2016) Amyloid hypothesis: Is there a role for antiamyloid treatment in late-life depression? Am J Geriatr Psychiatry 24, 239–247.

10.

Inada

, Inagaki

(2015) Psychotropic dose equivalence in Japan. Psychiatry Clin Neurosci 69, 440–447.

11.

American Psychiatric Association (2001) The Practice of Electroconvulsive Therapy: Recommendations for Treatment, Training and Privileging, 2nd Edition, American Psychiatric Press, Washington, DC.

12.

Yasuda

, Kobayashi

, Yamaguchi

, Tanaka

, Fujii

, Kitahara

, Tamaoki

, Matsushita

, Nunomura

, Motohashi

(2015) Seizure threshold and the half-age method in bilateral electroconvulsive therapy in Japanese patients. Psychiatry Clin Neurosci 69, 49–54.

13.

Baba

, Nakano

, Maeshima

, Satomura

, Kita

, Suzuki

, Arai

(2012) Metabolism of amyloid-β protein may be affected in depression. J Clin Psychiatry 73, 115–120.

14.

Nunomura

, Chiba

, Eto

, Saito

, Makino

, Miyagishi

(1996) Apolipoprotein E polymorphism and susceptibility to early- and late-onset sporadic Alzheimer’s disease in Hokkaido, the northern part of Japan. Neurosci Lett 206, 17–20.

15.

Petersen

, Aisen

, Beckett

, Donohue

, Gamst

, Harvey

, Jack

Jr , Jagust

, Shaw

, Toga

, Trojanowski

, Weiner

(2010) Alzheimer’s Disease Neuroimaging Initiative (ADNI): Clinical characterization. Neurology 74, 201–209.

16.

Hirata

, Matsuda

, Nemoto

, Ohnishi

, Hirao

, Yamashita

, Asada

, Iwabuchi

, Samejima

(2005) Voxel-based morphometry to discriminate early Alzheimer’s disease from controls. Neurosci Lett 382, 269–274.

17.

Matsuda

, Mizumura

, Nagao

, Ota

, Iizuka

, Nemoto

, Takemura

, Arai

, Homma

(2007) Automated discrimination between very early Alzheimer disease and controls using an easy Z-score imaging system for multicenter brain perfusion single-photon emission tomography. AJNR Am J Neuroradiol 28, 731–736.

18.

Direk

, Schrijvers

, de Bruijn

, Mirza

, Hofman

, Ikram

, Tiemeier

(2013) Plasma amyloid β, depression, and dementia in community-dwelling elderly. J Psychiatr Res 47, 479–485.

19.

Gurol

, Irizarry

, Smith

, Raju

, Diaz-Arrastia

, Bottiglieri

, Rosand

, Growdon

, Greenberg

(2006) Plasma β-amyloid and white matter lesions in AD, MCI, and cerebral amyloid angiopathy. Neurology 66, 23–29.

20.

Sutton

, Sohrabi

, Rainey-Smith

, Bird

, Weinborn

, Martins

(2015) The role of APOE-ɛ4 and beta amyloid in the differential rate of recovery from ECT: A review. Transl Psychiatry 5, e539.

21.

Hausner

, Damian

, Sartorius

, Frölich

(2011) Efficacy and cognitive side effects of electroconvulsive therapy (ECT) in depressed elderly inpatients with coexisting mild cognitive impairment or dementia. J Clin Psychiatry 72, 91–97.

22.

Zimmermann

, Schmitt

, Rotter

, Sperling

, Kornhuber

, Lewczuk

(2012) Transient increase of plasma concentrations of amyloid β peptides after electroconvulsive therapy. Brain Stimul 5, 25–29.

23.

Piccinni

, Veltri

, Vizzaccaro

, Catena Dell’Osso

, Medda

, Domenici

, Vanelli

, Cecchini

, Franceschini

, Conversano

, Marazziti

, Dell’Osso

(2013) Plasma amyloid-β levels in drug-resistant bipolar depressed patients receiving electroconvulsive therapy. Neuropsychobiology 67, 185–191.