Depression associated with dementia with Lewy bodies (DLB) and the effect of somatotherapy


  • This original article was presented by the author in Symposium of the 23rd annual meeting of Japanese Psychogeriatric Society in Kobe, 27–28 June 2008.

Dr Sho Takahashi MD, Clinical Neuroscience, Graduate School of Comprehensive Human Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba city, Ibaraki 305-8575, Japan. Email:


Background:  Dementia with Lewy bodies (DLB) is a common type of dementia. It is difficult to make an initial diagnosis of DLB because of a variety of early symptoms, including psychosis-like and depressive states. In this study, we examined the characteristic depressive symptoms of the prestage of DLB and the efficacy and safety of somatotherapy for depression accompanying DLB.

Methods:  Subjects in the study were 167 consecutive clinical cases aged 50 years or more, hospitalized at Tsukuba University Hospital from December 2002 to September 2007. At the time of admission, patients were diagnosed with certain types of mood disorders according to the Diagnostic and Statistical Manual of Mental Disorders Fourth Edition Text Revision. For each subject, a series of neuropsychological tests, along with a standard psychiatric and neurological assessment and biological examinations, were conducted. Using the data from these exams, we diagnosed probable and possible DLB according to the criteria for dementia with Lewy bodies established by McKeith et al.

1 We compared patients' depressive symptoms according to the Hamilton Depression Scale, and distinguished between patients with depression associated with DLB and those with other mood disorders.

2 We also examined the efficacy and safety of somatotherapy (electroconvulsive therapy (ECT) and transcranial magnetic stimulation (TMS)) for patients with drug therapy–resistant depression associated with DLB.

Results: 1 The characteristic symptoms of patients with DLB were classified into two groups: psychotic and non-psychotic. The former consisted of patients with states such as delusion and agitation, and the latter included patients exhibiting psychomotor retardation, loss of insight and hypochondriasis.

2 Eight DLB patients with therapy-resistant depression underwent ECT. After ECT, significant improvement was observed, with no remarkable safety hazards. Six patients with drug therapy–resistant DLB underwent TMS. TMS appears to be an effective, safe remedy for this kind of patient.

Conclusions: • A total of 13.8% of patients came to be re-diagnosed as having DLB as a consequence of a thorough examination after admission.

• Patients with depression associated with DLB were classified into psychotic and non-psychotic clusters.

• ECT and TMS are effective and safe therapeutic tools for drug therapy–resistant depression observed in DLB patients.


We sometimes encounter presenile or elderly patients with a mood disorder who subsequently develop dementia. They are often referred to us under the diagnosis of refractory depression also showing a variety of adverse effects related to psychotropics.

Dementia with Lewy bodies (DLB) has been established as one of the three major types of dementia, along with Alzheimer's disease (AD) and vascular dementia. According to the diagnostic criteria for DLB, in principle, the typical clinical symptoms include progressive, reduced cognitive function, variable, visual hallucinations and Parkinsonism.1 It has been said that more than half of DLB patients are diagnosed with depression during their clinical course. Previous researchers have reported that the association between depression and DLB is stronger than that of other dementia causing illnesses. For example, McKeith et al. found that the prevalence of depressive symptoms was higher in patients with DLB than in patients with AD.2

Some patients develop depression as a prodromal, early symptom and subsequently start to show the symptoms that fulfill the DLB criteria. As early as the 1980s, Reding et al. reported that patients with depression who also showed some of the following characteristic symptoms were likely to develop dementia in the future: the manifestation of a confused state after the administration of low-dose tricyclic antidepressants and the presence of extrapyramidal signs. This report seems to indicate the relationship between DLB and depression. In addition, a study examining patients with so-called pseudo-dementia reported that five of 16 patients with dementia secondary to depression were subsequently diagnosed with Parkinson's disease plus dementia.3 Therefore, accurate early diagnosis may improve the prognosis for such a patient, as well as their quality of life.

It is well known that DLB patients often show hypersensitivity to psychotropics, including antipsychotics. DLB patients also occasionally develop autonomic nervous system (such as respiratory/circulatory system) dysfunction. Taken together, electroconvulsive therapy (mECT, ECT) and repetitive transcranial magnetic stimulation (rTMS) appear to be preferable to pharmacotherapy for patients with DLB. Thus, in this study we also evaluated the efficacy and safety of these treatments.


Characteristic depressive symptoms observed in DLB patients

Subjects in the study included 167 consecutive clinical cases aged 50 years or more who were hospitalized in the psychiatric ward of Tsukuba University Hospital between December 2002 and September 2007. They were diagnosed as having mood disorders according to the Diagnostic and Statistical Manual of Mental Disorders Fourth Edition Text Revision (DSM-IV-TR) at the time of admission, specifically, either major depressive disorder or bipolar I disorder. Our study was confined to a group of patients who completed a series of examinations, as described below. The ‘revised criteria for the clinical diagnosis of dementia with Lewy bodies’1 was used to make a diagnosis of probable and possible DLB. We also used the results of the examinations, which included laboratory testing (hematological, serum chemistry), neuroimaging (magnetic resonance imaging and single-photon emission computed tomography (SPECT)), neuropsychological testing and an examination of autonomic function including hypercapnic ventilation response (HVR).4–6

There were a number of patients among the target population who met the following conditions: a score of 24 or more on the mini-mental state examination (MMSE) scale, not meeting the criteria of dementia specified in DSM-IV-TR, and satisfying some of the criteria for DLB but falling short the diagnosis of DLB. We defined such patients as suspected DLB with supportive features (hereinafter referred to as ‘suspected DLB’). We then classified patients into three types: probable DLB, possible DLB and suspected DLB. We estimated the prevalence of these three types among the 167 patients. We compared depressive symptoms using the Hamilton Depression Scale (HAM-D) between the pooled subjects of the three DLB groups and patients with other types of depression. For the comparison, we made matches regarding the HAM-D score, along with age and gender between the two groups. The Mann–Whitney U-test was used for statistical analyses.

Somatotherapy for depression in DLB patients

Eight patients (one male, seven females, with a mean age (± SD) of 71.6 (± 7.3) years) with a diagnosis of DLB and drug therapy–resistant depression underwent ECT. Conventionally, therapy-resistant depression has been defined as a pathological condition in which the patient does not respond to more than two antidepressants with different mechanisms of action.3 We also adopted this definition for our study. Three patients were diagnosed with possible DLB and five with probable DLB.

Six patients (three males, three females, with a mean age (± SD) of 61.9 (± 9.2) years) with DLB underwent TMS. (Five were diagnosed as suspected DLB and one as probable DLB.)

After obtaining informed consent from each patient, ECTs were conducted through electrodes positioned at the standard bifrontotemporal location. For pulse wave stimuli, a Thymatron System IV ECT apparatus (Somatics, Lake Bluff, IL, USA) containing an inbuilt electroencephalography system (Fp1-A 1, Fp2- A2, international 10–20 system) was used. Stimulation dose was calculated using the ‘half age’ method. A LOW 0.5 preset program using 0.5 ms pulse width was selected, adjusting frequency to maximize duration. The criterion for an adequate seizure was an electroencephalographic seizure lasting 20 s. If no electroencephalographic seizure had occurred after 20 s, re-stimulation at a higher stimulus intensity was immediately performed by increasing voltage by 5∼10% for pulse wave stimuli, to a maximum of two stimulations/session.7 Motor seizures were further monitored in a cuffed arm. Intravenous injection of thiamylal sodium and succinylcholine was performed for patients given pulse wave stimuli. Antidepressants remained unchanged at a minimal dose throughout the course of ECT. Lithium carbonate and sodium valproate were withdrawn before first ECT. A treatment course consisted of six energizations.

We used the Magstim Rapid System (MRS 1000/50) (Magistim Company, Carmarthenshire, UK) as the stimulator for rTMS and selected a 70-mm figure-of-eight coil.8 We stimulated the dorsolateral prefrontal cortex area, approximately 5 cm ahead of the site where the maximum exercise-induced reaction could be obtained, on the right and left side. The stimulus intensity was adjusted to 110% motor threshold on the right side and 100% motor threshold on the left side. A course consisted of the following treatments. A train of stimuli (1 Hz × 140 s (140 pulses)) was administered on the right side at intervals of 30 s three times a day (420 pulses/day). A train of stimuli (10 Hz × 5 s (50 pulses)) was administered on the left side at intervals of 25 s 15 times a day (750 pulses/day).8,9 These treatments were repeated for 10 days.

We evaluated the patients' depressive symptoms before and after the ECT and TMS treatment sessions using the HAM-D, and compared the results.


Dementia with Lewy bodies (DLB) and depression

Length of hospitalization, age and gender difference of subjects

The 167 patients were hospitalized for 71 ± 48 days. Their mean age was 63 ± 9 years, ranging from 50 to 83 years. According to the above-described diagnostic procedure, 23 patients were classified into the three DLB groups (13.8%), while the remaining 144 patients were classified into the non-DLB group (86.2%). The male-to-female ratio was 26% vs 74% for the DLB group and 36% vs 64% for the non-DLB group. The ratio was higher for females in the two groups. The mean age was 63.5 ± 9.2 years in the DLB group and 63.2 ± 9.0 years in the non-DLB group.

In our ward, a tentative psychiatric diagnosis for each patient is made on the consensus of two psychiatrists using the DSM-IV-TR at the time of admission. Applying this procedure to our study, of the 23 DLB patients, 22 were diagnosed with major depressive disorder (95.7%) and one with bipolar I disorder at the time of admission (Table 1).

Table 1.  Diagnosis of depressive disorders for each type of DLB
  1. DLB, dementia with Lewy bodies.

Major depressive disorder9103
Bipolar disorder001
(Bipolar I disorder)
Dysthymic disorder000
Mood disorder resulting from general medical condition000

The results of a series of examinations are shown in Table 2. As shown, 50% of the patients with suspected DLB were positive in 123I-metaiodobenzylguanidine (MIBG) myocardial scintigraphy. Higher positive findings were also obtained in a reduction in regional cerebral blood flow revealed by brain perfusion SPECT (88.9%), which is a characteristic finding of neuroimaging for DLB,10 along with a poor hypercapnic ventilation response (83.3%) (Table 2).

Table 2.  Evaluation in the group of dementia with Lewy bodies (DLB) with the initial episode of mood disorder
 Number of patients who tested positive/number of patients who underwent examination (%)
Suspected DLBPossibleProbable
  1. DLB, dementia with Lewy bodies; MIBG, 123I-metaiodobenzylguanidine myocardial scintigraphy; r-CBF, regional cerebral blood flow; SPECT, single-photon emission computed tomography.

(Reduced r-CBF in occipital lobe)
Ventilation response5/65/60/0

Age, gender and MMSE scores of the patients with the three types of DLB

The group diagnosed as suspected DLB with supportive features consisted of nine patients (one male, eight female) with a mean age of 59.3 ± 10.4 years. The group diagnosed as possible DLB consisted of 10 patients (three male, seven female) with a mean age of 70.1 ± 6.5 years. The group diagnosed as probable DLB included four patients (two male, two female) with a mean age of 69.0 ± 11.3 years. The mean MMSE score was 19.7 ± 3.0 for the possible DLB group, 21.7 ± 1.2 for the probable DLB group and 27.8 ± 1.9 for the suspected DLB group, respectively. The mean score was significantly higher for the suspected DLB group (Table 3).

Table 3.  Three types of DLB with an initial episode of mood disorder
  • *

    The patients in the suspected dementia with Lewy bodies (DLB) subgroup were significantly younger than those in the remaining two subgroups (P < 0.005). MMSE, mini-mental state examination.

n (23/167)9104
Mean age (± SD)59.3 ± 10.470.1 ± 6.5*69.0 ± 11.3*
MMSE27.8 ± 1.919.7 ± 3.021.7 ± 1.2

Characteristic symptoms of depression shown in DLB patients

Our analysis revealed that the scores for several symptoms were higher for the DLB group. They are divided into two clusters: psychotic (agitation, paranoia, depersonalization and derealization) and non-psychotic (psychomotor retardation, loss of insight, hypochondriasis). In general, each patient had either psychotic or non-psychotic symptoms, and the two clusters seldom coexisted in a study subject (Table 4).

Table 4.  Clinical signs of patients with depression accompanying DLB (ratings of depression using the 24-item HAM-D)
  1. Mann–Whitney U-test. DLB, dementia with Lewy bodies; HAM-D, Hamilton Depression Scale.

Symptoms observed with significant frequency in the three DLB groups were divided into the following two categories:
(1) Psychotic symptoms 
 Agitation(P = 0.002)
 Paranoid symptoms(P = 0.001)
 Depersonalization and derealization(P = 0.023)
(2) Non-psychotic symptoms 
 Psychomotor retardation(P = 0.022)
 Hypochondriasis(P = 0.045)
 Lack of insight(P = 0.007)


Electroconvulsive therapy (ECT)

As shown in Figure 1, the HAM-D score decreased from 38.0 ± 5.8 before ECT to 15.0 ± 9.6 after ECT, a difference that indicates significant improvement (P < 0.005).

Figure 1.

Changes in Hamilton Depression Scale (HAM-D) scores (electroconvulsive therapy (ECT)).

Transcranial magnetic stimulation (TMS)

Figure 2 shows that the HAM-D score decreased from 24.0 ± 8.0 before TMS to 11.0 ± 5.9 after TMS, a difference that also indicates significant improvement (P < 0.005).

Figure 2.

Changes in Hamilton Depression Scale (HAM-D) scores (transcranial magnetic stimulation (TMS)).

Both ECT and TMS posed no safety hazard to the patients.


Dementia with Lewy bodies (DLB) and depression

As shown in the results section, approximately 14% of the presenile and senile patients who had been diagnosed as having depression or related disorders before or at the time of admission to our ward came to be re-diagnosed as having DLB at the time of their discharge. For the 10 possible and four probable DLB patients, each diagnosis was made according to the revised diagnosis criteria of DLB including a principal condition of presence of dementia. However, the diagnosis of suspected DLB, which is newly defined by us, is principally based on some clinical features listed on the criteria exclusive of the presence of dementia. With the aid of SPECT, HVR and MIBG myocardial scintigraphy data, we were able to make this diagnosis.

Needless to say, we do not mean to imply that patients with suspected DLB must progress to possible or probable DLB in the future. However, we reported elsewhere5,6 that no positive findings were observed in SPECT, HVR and scintigraphy data among the healthy elderly. In addition, it has been reported that the mean value for the sensitivity of the former ‘criteria for the clinical diagnosis of DLB’11 was 49%.12 This value is not so good, but this criteria alone can make a diagnosis of DLB to a certain degree. Because we used the revised version of the criteria,1 sensitivity is expected to be higher than that for the former. Taken together, we can assume a certain portion of the patients with suspected DLB might convert to possible or probable DLB in the future. In conclusion, on the examination of presenile and senile patients with refractory depression, we must keep in mind the diagnosis of DLB.

Characteristics of depressive symptoms

It is well known that patients with DLB develop psychosis. Kosaka described as follows, ‘DLB often starts with psychosis. Before DLB came to be acknowledged as a common pathological condition, not a few patients with DLB had been misdiagnosed as having schizophrenia and received inappropriate treatment, and finally diagnosed as having DLB after autopsy.’13 In fact, we found a group of patients with a psychosis-like state who had been occasionally diagnosed as having Cotard syndrome.

Regarding this issue, Aarsland et al. examined the psychiatric symptoms associated with Parkinson's disease with dementia (PDD) among 537 patients with PDD using the neuropsychiatric inventory. They classified the symptoms into five clusters: few and mild symptoms (52%), mood (11%), apathy (24%), agitation (5%) and psychosis (8%).

As shown above, we analyzed HAM-D subscale scores and found that the DLB group consisted of the following two subgroups: non-psychotic and psychotic. The former is characterized by apathy and paucity of sadness, the latter by prominent psychosis. According to traditional depression diagnoses, patients of this group can be diagnosed as having Cotard syndrome or agitated depression. Although DLB and PDD are distinguishable clinical entities, the two clusters proposed by Aarsland et al. (agitation (high score on agitation and high total neuropsychiatric inventory score); psychosis cluster (high scores on delusions and hallucinations)) might correspond to our psychotic group.14 Our non-psychotic group might correspond to their mood or apathy clusters. Therefore, we should recall DLB when presenile or senile patients with a depressive state show marked apathy or psychotic features.

Somatotherapy for DLB

Electroconvulsive (ECT) and TMS therapy

We made a comparison between the ECT group and the TMS group with respect to clinical variables. According to the results, it appears that ECT was selected as a last resort for patients with DLB who showed the following characteristics: severe depression, poor response to various treatments, and the requirement of urgent psychiatric and physical intervention. On the contrary, TMS seems to have been employed for the patients with a milder form of depression. Regarding the efficacy, both ECT and TMS appear to be useful therapeutic options for drug therapy-resistant DLB.

Adverse reactions and safety of ECT and TMS

Electroconvulsive (ECT) entails the risks of critical circulatory and respiratory accidents, as well as headache and defect of memory.15 As for TMS, it has been said that 10–20% of patients develop headache as an adverse reaction of rTMS.16 Compared with ECT, rTMS has been reported to entail fewer risks of critical accidents.

Although both ECT and TMS posed no safety hazards in this study, many of the DLB patients had dysfunction of the autonomic nervous system, including the circulatory and respiratory systems. A thorough pretreatment examination is indispensable before the implementation of ECT for DLB patients.


  • • Of the presenile and senile patients who were hospitalized with a diagnosis of mood disorder, 13.8% came to be re-diagnosed as having DLB as a consequence of a thorough examination after admission.
  • • The patients with depression associated with DLB were classified into the following two clusters: a psychotic cluster characterized by delusion and agitation and a non-psychotic cluster characterized by psychomotor retardation, loss of insight and hypochondriasis.
  • • ECT and TMS are effective and safe therapeutic tools for drug therapy–resistant depression observed in DLB patients.