Introduction: Lewy body disease (LBD) is the second most common neurodegenerative disorder in patients older than 65 years. LBD is characterized by heterogeneous symptoms like fluctuation in attention, visual hallucinations, Parkinsonism, and REM sleep behaviour disorders. Considering the relevant social impact of the disease, identifying effective non-pharmacological treatments is becoming a priority. The aim of this systematic review was to provide an up-to-date literature review of the most effective non-pharmacological treatments in patients with LBD, focussing on evidence-based interventions. Methods: Following PRISMA criteria, we carried out a systematic search through three databases (PubMed, Cochrane Libraries, and PEDro) including physical therapy (PT), cognitive rehabilitation (CR), light therapy (LT), transcranial direct current stimulation (tDCS), transcranial magnetic stimulation (TMS), electroconvulsive therapy (ECT), deep brain stimulation (DBS). All studies were qualitatively assessed using standardized tools (CARE and EPHPP). Results: We obtained a total of 1,220 studies of which 23 original articles met eligibility criteria for inclusion. The total number of LBD patients included was 231; mean age was 69.98, predominantly men (68%). Some PT studies highlighted improvements in motor deficits. CR produced significant improvements in mood, cognition, and patient’s quality of life and satisfaction. LT outlined a partial trend of improvements in mood and sleep quality. DBS, ECT, and TMS showed some partial improvements mainly on neuropsychiatric symptoms, whereas tDCS provided partial improvements in attention. Conclusion: This review highlights the efficacy of some evidence-based rehabilitation studies in LBD; however, further randomized controlled trials with larger samples are needed to provide definitive recommendations.

© 2023 S. Karger AG, Basel

Introduction

Lewy body disease (LBD) is a neurodegenerative disorder characterized by a clinical spectrum in which dementia with Lewy bodies (DLB) and Parkinson’s disease dementia (PDD) represent different points of the disease continuum [1, 2]. Both DLB and PDD share common clinical manifestations like motor symptoms, cognitive impairments (mainly attentional, executive, and visuoperceptual deficits), sleep disorders, and neuropsychiatric symptoms [3]. Despite these similarities, the timing of the clinical presentation could differentiate the two syndromes according to the 1-year rule recommended by the international consensus of the Dementia with Lewy Bodies Consortium. DLB should be diagnosed when dementia occurs before or concurrently with Parkinsonism, whereas PDD should be diagnosed when dementia occurs in the context of well-established Parkinson’s disease [4].

The prevalence rate of LBD is considerable, being the second most common neurodegenerative disease in older people, although diagnoses of DLB and PDD are often delayed and could be under-recognized [5, 6]. Clinical manifestations of LBD may be very heterogeneous in terms of symptomatology and timing of presentation across individuals, making the treatment management difficult. For instance, DLB and PDD might respond differently to the same treatments, underlining the importance of an accurate diagnosis [3]. Furthermore, pharmacological interventions may present the intrinsic risks of improving one symptom but worsening another (e.g., a pharmacological intervention addressing neuropsychiatric symptoms may exacerbate motor symptoms and vice versa) [7].

In this perspective, identifying effective non-pharmacological treatments to slow down the worsening of the disease, thus improving both patients and caregivers’ quality of life should be a priority. In recent years, there have been a growing number of non-pharmacological trials involving LBD patients, although no definitive recommendations are provided. In a previous systematic review, Inskip and colleagues [8] collected all physical therapy (PT) interventions for LBD patients presented in the literature, and they concluded that these treatments showed some improvements in gait speed, although the quality of the studies was low and with a restricted number of participants. Other two systematic reviews focused on non-pharmacological interventions in LBD and presented a very heterogeneous range of treatments [9, 10]. Both reviews identified possible benefits of non-pharmacological interventions, although the studies highlighted several limitations due to the small sample sizes and the low quality of the study designs. The aim of the present systematic review was to provide an up-to-date literature review of the most effective non-pharmacological treatments in patients with Lewy body disease, focussing on evidence-based interventions.

Methods

The systematic review was conducted according to the PRISMA guidelines [11].

Eligibility Criteria

For the selection of the studies, PICOS inclusion criteria were followed: population, intervention, comparison, outcome measures, and study design. Only original articles written in English language were included.

The target population was patients with a diagnosis of DLB [4] or PDD [12], both pathologies included under the umbrella term Lewy body disease. Therefore, individuals with different age, gender, and ethnicity with a clinical diagnosis of DLB or PDD were included. Studies that explored the efficacy of non-pharmacological interventions without separately reporting outcomes for LBD were not included. For the purpose of the present review, we included non-pharmacological interventions concerning the following domains: PT, cognitive rehabilitation (CR), light therapy (LT), transcranial direct current stimulation (tDCS) and transcranial magnetic stimulation (TMS), electroconvulsive therapy (ECT), deep brain stimulation (DBS). Studies reporting any other non-evidence-based treatments, such as musical therapy or psychoeducational interventions, were not included. All selected studies contained quantitative and/or qualitative outcome measures as indicators of the treatment’s efficacy.

Search Strategy

A systematic search was conducted in March 2022 through the following three free databases: PubMed, Cochrane libraries, and PEDro (Physiotherapy Evidence Database). All selected terms (MeSH terms) were written combining appropriate syntax for each database (see online suppl. material at www.karger.com/doi/10.1159/000529256). No filters were applied to the search nor to the records extracted from databases.

Study Selection

Two independent reviewers (L.G. and M.M.) performed separately the study selection to guarantee the consistency of the results. The first step, after extracting all records from the three databases, was to remove all duplicates using Python software (https://www.python.org/). The second step was reading the titles and the abstracts of the studies and excluding those presenting unrelated topics. After this preliminary screening process, a final pool of articles was read in full and those that followed the eligibility criteria were included in the review.

Quality Assessment

The quality of the studies included was also assessed independently by two reviewers (LG and MM) using standardized tools for case reporting: “CARE criteria checklist” [13] and the “Effective Public Health Practice Project Quality Assessment Tool for Quantitative Studies” [14]. Any disagreements between reviewers were resolved through discussion.

ResultsStudy Selection

The systematic search identified a total of 1,220 records from the three databases (PubMed, Cochrane Libraries, and PEDro) of which 1,180 were unique and 40 duplicates. Other 10 additional studies were identified through the reference list of the selected papers and previous reviews on this topic. After excluding articles unrelated to our topic, 29 studies were read in full, and a final number of 23 studies met the eligibility criteria, shown in the flowchart (see Fig. 1). The included studies were 8 randomized controlled trials, 7 uncontrolled trials, and 8 case studies.

Fig. 1.

PRISMA flow diagram of the systematic review.

Quality Assessment

All following trials (n = 15) were evaluated with EPHPP quality assessment tool [14]: three received strong global rating, other three a moderate global rating, and the majority (n = 9) received a weak global rating. Most frequently, these studies failed to control for confounders (n = 8), blinding condition (n = 7) and others do not provide a control group (n = 10). All case studies (n = 8), evaluated with CARE checklist [13], did not provide a patient’s history organized as a timeline. Furthermore, due to the intrinsic nature of case studies, some biases are clearly present (e.g., selection bias, confounders, and blinding).

Participants

The total number of participants included in the present review was 231. Of these, 113 patients were reported with a diagnosis of DLB, 103 with PDD, and 15 described as LBD. The mean age of the participants was 69.98 years (range 54.4–94) and although information about gender was not reported in 1 study, participants were clearly predominantly men (68%) (Table 1). Cognitive profiles were tested at baseline in most cases, and participants varied from those with severe dementia [15] to very mild dementia [16]. Information about medications was present in almost all studies (n = 19/23 studies), and drug intake remained stable for the entire duration of the treatments except for ECT (Table 1). Patients were recruited from neurological and movement disorders clinics (n = 7 studies), psychiatric departments (n = 4 studies), general hospitals (n = 2 studies), nursing home (n = 1 study), and n = 9 studies did not report patients’ provenience.

Table 1.

Characteristics of LBD patients included in this systematic review

CitationStudy designNumber of participantsAge of participants mean (SD)GenderReported diagnosisMedications during treatmentPT Tabak et al. [20] 2013Case study1611 MPDDCarbidopa + levodopa Dawley et al. [19] 2015Case study1571 MLBDCarbidopa/levodopa; antidepressant; antipsychotic Telenius et al. [18] 2015Randomized controlled trial484 (10)1 M; 3 FPDDNR Longhurst et al. [17] 2020Uncontrolled trial3577 (6)24 M; 11 FDLBNR Kegelmeyer et al. [21] 2021Uncontrolled trial883.57 (6.58)4 M; 4 FLBDNRCR Hindle et al. [22] 2018Randomized controlled trial2976.34 (6.42)23 M; 6 F25 PDD; 4 DLBLevodopa (all patients) + cholinesterase inhibitors and N‐methyl‐D‐aspartate receptor antagonists (9/29)LT Sekiguchi et al. [15] 2017Uncontrolled trial574.4 (7.3)3 M; 2 FDLBAripiprazole (1/5); tiapride (1/5); sodium valproate (1/5); rivastigmine (1/5); galantamine (1/5); quietapine (3/5) Akkaoui et al. [23] 2019Case study1631 MDLBVerapamiltDCS and TMS Elder et al. [26] 2016 tDCSUncontrolled trial1364.81 (7.9)10 M; 3 F8 PDD; 5 DLBCholinesterase inhibitors (1/13); antidepressants (3/13); levodopa (all patients) Elder et al. [27] 2017 tDCSRandomized controlled trial3866.63 (8.39)27 M; 11 FPDDCholinesterase inhibitors + levodopa (all patients) Elder et al. [25] 2019 tDCSRandomized controlled trial3675.09 (7.97)27 M; 9 F13 PDD; 23 DLBLevodopa + cholinesterase inhibitors (all patients) Wang et al. [24] 2021 tDCSRandomized controlled trial1177.08NRDLBNR Takahashi et al. [28] 2009a TMSUncontrolled trial661.9 (9.2)3 M; 3 FDLBNRECT Kung et al. [31] 2002Case study1601 FDLBSertraline + citalopram Rasmussen et al. [29] 2003Uncontrolled trial773.57 (10.56)2 M; 5 FDLBAntidepressants, antipsychotics, mood stabilisers (all patients), cholinesterase inhibitors (5/7); carbidopa/levodopa (2/7) Takahashi et al. [28] 2009aUncontrolled trial871.6 (7.3)1 M; 7 FDLBAntidepressant Izuhara et al. [30] 2020Case study1691 FDLBSuvorexantDBS Loher et al. [39] 2002Case study1751 MPDDLevodopa + carbidopa + paroxetine Freund et al. [36] 2009Case study1711 MPDDLevodopa Ricciardi et al. [16] 2015Case study1691 MPDDLevodopa Kim et al. [35] 2017Uncontrolled trial566 (1.8)3 M; 2 FPDDLevodopa (all patients) Gratwicke et al. [33] 2018Randomized controlled trial665.2 (10.7)6 MPDDCholinesterase inhibitors + levodopa (all patients) Gratwicke et al. [34] 2020Randomized controlled trial671.33; (3.67)5 M; 1 FDLBCholinesterase inhibitors (all patients) + levodopa (5/6) Maltête et al. [32] 2021Randomized controlled trial662.2 (7.8)6 MLBDCholinesterase inhibitors (all patients)PT Studies

Primary outcomes mainly used to evaluate the efficacy of physical therapies were balance, gait, and cognitive performance (Table 2). Significant improvements in all motor measures were found in a recent retrospective study in which 35 DLB participants underwent a 4-week structured PT programme in a clinical setting [17]. In a randomized controlled trial [18], involving a large cohort of persons with dementia, the following measures relative to 4 PDD participants were reported separately: after 12 weeks of intensive strengthening and balance exercises, 2 PDD assigned to the exercise group improved sit-to-stand, balance, and speed measures compared to the 2 PDD assigned to the control group. In a case study, Dawley [19] reported a relatively young patient (age 57) with a diagnosis of LBD treated with a Parkinson’s specific intervention called “Lee Silverman Voice Treatment-Big,” a programme with intensive exercises with large amplitude movements of the body. The patient after 12 weeks of training improved in all motor scales (sit to stand, speed, risk of falls, walking, and balance), although results were not statistically significant due to the nature of the study. Another case study by Tabak and colleagues [20] reported successful 8-week PT in a PDD patient that obtained benefits in both motor functions (walking and balance) and executive functions, measured with Montreal Cognitive Assessment subtests, suggesting a causal relationship between these abilities. Finally, Kegelmeyer and colleagues [21], in a cohort study of 8 LBD patients, administered a single session of treadmill training with the goal of improving gait disorders. No significant changes in gait measures were found after a session of treadmill walking, and authors highlight the need of longer sessions to obtain significant improvements (Table 3).

Table 2.

Treatment characteristics of the included studies

CitationTreatmentCharacteristics of the treatmentDuration of treatmentPrimary outcomesSecondary outcomesTabak et al. [20] 2013PTAerobic exercise training on a stationary bicycle8 weeksEffects on executive functionsEffects on disease severity, QoL, walkingDawley et al. [19] 2015PTIntensive exercises (LSVT BIG programme)12 weeksEfficacy of a PD treatment with a LBD patient–Telenius et al. [18] 2015PTIntensive strengthening, balance exercises12 weeksEffects on balanceEffects on muscle strength, mobility, ADL, QoL, and neuropsychiatric symptomsLonghurst et al. [17] 2020PTAerobic activity, strengthening, balance training4 weeksEffects on gait and balanceEffects on cognitionKegelmeyer et al. [21] 2021PT20 min of treadmill training1 dayFeasibility and safety of the treatmentEffects on gait, mobility, and coordinationHindle et al. [22] 2018CROrientation, planning, and memory exercises8 weeksEffects on goal attainment and satisfactionEffects on cognition, mood, QoLSekiguchi et al. [15] 2017LT2,500–5,000 lux2 weeksEffects on different type of dementiaEffects on different severity of dementiaAkkaoui et al. [23] 2019LT10,000 lux6 weeksEffects on sleep disturbances–Elder et al. [26]2016tDCSAnodic stimulation in left DLPFC1 dayFeasibility of the treatment and the effects on attention–Elder et al. [27] 2017tDCSAnodic stimulation in left DLPFC1 dayEffects on attention–Elder et al. [25] 2019tDCSAnodic stimulation in right posterior parietal cortex4 daysEffects on frequency and severity of visual hallucinationsEffects on visual cortical excitability and visuoperceptual functionWang et al. [24] 2021tDCSAnodic stimulation in left DLPFC10 daysEffects on cognition–Takahashi et al. [28] 2009aTMSBilateral DLPFC10 daysEfficacy and safety of TMS in depressive patients–Kung et al. [31] 2002ECTUnilateral7 sessionsEffects on mood and neuropsychiatric symptoms–Rasmussen et al., [29] 2003ECTBitemporal stimulation (depressive symptoms)
Bifrontal stimulations (cognitive symptoms)Different treatments (4–33 sessions)Effects on mood–Takahashi et al. [28] 2009aECTBifrontotemporal10 sessionsEfficacy and safety of ECT in depressive patients–Izuhara et al. [30] 2020ECTBitemporal15 sessionsEffects on psychiatric symptoms and cognitive fluctuations–Loher et al. [39] 2002DBSLeft internal segment of the globus pallidus stimulation>1 yearEffects on disabling motor fluctuations and severe dyskinesia–Freund et al. [36] 2009DBSHigh frequency (130 Hz) of bilateral STN and low frequency (20 Hz) of bilateral NBM stimulation23 weeks (STN) −18 weeks (NBM)Effects on cognition–Ricciardi et al. [16] 2015DBS30 Hz frequency of unilateral PPN stimulation4 yearsEffects on cognition–Kim et al. [35] 2017DBSBilateral STN stimulation4–8 yearsEffects on motor and non-motor symptoms–Gratwicke et al. [33] 2018DBS20 Hz frequency of bilateral NBM stimulation6 weeksEffects on cognitionEffects on psychiatric, motor symptoms and fMRI resting stateGratwicke et al. [34] 2020DBS20 Hz frequency of bilateral NBM stimulation6 weeksSafety and tolerability of NBM DBS procedureEffects on cognitive, psychiatric, motor scales and functional connectivityMaltête et al. [32] 2021DBS20–100 Hz frequency of bilateral NBM stimulation3 monthsEffects on a memory test score (FCSRT)Safety and effects on cognition, motor deficits, sleep, and PETTable 3.

Summary of the results of physical therapy (PT) studies

CitationTarget of the treatmentMeasureBaseline scores mean (SD)Scores after treatment mean (SD)Tabak et al. [20] 2013CognitionMOCA1724MotorUsual gait speed, m/s0.960.92Walking (2MWT)100129Balance (FGA)1323Dawley et al. [19] 2015MotorSit to stand (CST)48Risk of falls (TUG)15.459.05Usual gait speed m/s0.81.43Walking (6MWT)480562Balance (MBT)2125Telenius et al. [18] 2015Case 1CognitionMMSE16NRMotorSit to stand (CST)68Usual gait speed, m/s0.350.3Balance (BBS)2327Case 2CognitionMMSE16NRMotorSit to stand (CST)58Usual gait speed m/s0.410.71Longhurst et al. [17] 2020CognitionMOCA16.9 (6.7)17.8 (6.9)MotorSit to stand (5STS)19.3 (13.5)14.7 (5.6)aRisk of falls (TUG)13.5 (10.2)11.1 (6.7)aUsual gait speed, m/s0.90 (0.27)1 (0.27)aWalking (6MWT)348 (105)381 (120)aBalance (MBT)18.2 (4.4)20.4 (4.5)aKegelmeyer et al. [21] 2021CognitionMMSE19.14 (10.83)NRMotorRisk of falls (TUG)24.56 (14.99)21.57 (12.12)Usual Gait speed, m/s0.66 (0.25)0.79 (0.23)CR Study

Our systematic search identified only one intervention [22] of CR for a group of 29 LBD patients that were randomly assigned to the following 3 groups: cognitive rehabilitation (CR), relaxation therapy, and treatment as usual. Primary outcomes of the study were goal attainment and satisfaction of the intervention, and secondary outcomes were quality of life, cognition, and mood (Table 2). After 8 weeks of CR consisting of planning, orientation, and memory exercises, participants assigned to CR group compared to relaxation therapy and treatment as usual groups were less depressed, more socially involved, with higher perception of self-efficacy and with higher values in goal attainment and satisfaction. After 6 months at the follow-up, the CR group showed again higher values in goal attainment, higher rating in questionnaires assessing quality of life (The Parkinson’s Disease Questionnaire; the Euroqol Questionnaire‐short version), and better cognitive performance in a memory test (Table 4).

Table 4.

Summary of the results of cognitive rehabilitation (CR)

CitationTarget of the treatmentMeasureBaseline scores mean (SD)After treatment scores (2 months) mean (SD)Follow-up scores (6 months) mean (SD)Hindle et al. [22] 2018Goal AttainmentBGSI3.08 (1.43)6.29 (1.44)a (CR vs. TAU and RT)6.6 (1.93)a (CR vs. TAU and RT)SatisfactionBGSI3.3 (1.36)6.54 (1.48)a (CR vs. TAU and RT)5.98 (1.7)Neuropsychiatric symptomsDepression (HADS)9.13 (1)5.5 (3.5)a (CR vs. TAU)6.14 (4.14)Quality of lifePhysical (WHOQOL‐BREF)13.2 (2.57)12.5 (3.12)13.15 (2.1)Psychological (WHOQOL‐BREF)14.5 (2.22)13.25 (2.82)14.49 (2.65)Social (WHOQOL‐BREF)14.4 (3.92)15.85 (2.31)a (CR vs. TAU and RT)15.47 (2.02)Environmental (WHOQOL‐BREF)15.7 (2.54)16.13 (2.23)15.98 (1.49)PDQ821.56 (15.27)29.3 (10.95)26.18 (16.1)a (CR vs. TAU)ED5D3L0.65 (0.27)Not measured0.59 (0.31)a (CR vs. TAU)GSES31 (4.15)31.5 (4.24)a (CR vs. RT)31.83 (5.07)CognitionMemory recall (RBMT)1.7 (2.46)2.17 (1.36)3.06 (1.57)a (CR vs. TAU)Attention (TMT)4.33 (3.21)4 (3.46)4.5 (2.1)Verbal fluency27.4 (11.83)30.13 (14.4)23.14 (7.58)Functional activityFAQ9.5 (7.04)Not measured13.57 (7.87)LT Studies

Only two studies investigated the efficacy of LT on both sleep quality and neuropsychiatric symptoms in LBD patients. Sekiguchi and colleagues [15] presented a case series of 5 DLB patients that underwent 2 weeks of daily light treatment with 2,500–5,000 lux; none of the patients improved in sleep score after treatment. Instead, in a case study, Akkaoui and colleagues [23] showed in a DLB patient clear improvements in both depression and sleep measures after 6 weeks of daily treatment with 10,000 lux, although the nature of the study (single case) does not allow statistically significant results (Table 5).

Table 5.

Summary of the results of light therapy (LT) studies

CitationTarget of the treatmentMeasureBaseline scoresAfter treatment scoresSekiguchi et al. [15] 2017Case 1Neuropsychiatric symptomsNPI (sleep score)83Case 2Neuropsychiatric symptomsNPI (sleep score)1212Case 3Neuropsychiatric symptomsNPI (sleep score)99Case 4Neuropsychiatric symptomsNPI (sleep score)86Case 5Neuropsychiatric symptomsNPI (sleep score)88Akkaoui et al. [23] 2019Neuropsychiatric symptomsNPI (sleep score)107Depression (MADRS)2211SleepPSQI2213ESS1712tDCS and TMS Studies

Interventions with tDCS were generally addressed to ameliorate cognition (especially attention) and neuropsychiatric symptoms (especially hallucinations). A randomized controlled trial by Wang and colleagues [24] tested the effect of 10 consecutive sessions of tDCS over the left dorsolateral prefrontal cortex (DLPFC) in 11 DLB patients. Neuropsychological evaluation, pre- and post-rehabilitation, did not show any difference between the active and sham groups. Another randomized controlled trial involving 36 LBD patients who underwent 4 consecutive sessions of tDCS with a focus on the right posterior parietal cortex did not report a reduction in the frequency and severity of visual hallucinations [25]. Elder and colleagues [26, 27] also tried to explore the effect of a single session of tDCS over the left DLPFC: one trial was effective in improving attentive measures like choice reaction time and digit vigilance [26], and the other did not report any beneficial effects on LBD participants [27]. In addition, there was a single trial in which 6 DLB participants underwent 10-day sessions of TMS. The case series by Takahashi and colleagues [28] showed that stimulating the left and right DLPFC with the TMS reduced significantly depressive symptoms (HAM-D score before treatment = 24; HAM-D after treatment = 11) (Table 6). In the same study, the authors treated another group of LBD patients with an electroconvulsive treatment (Table 7).

Table 6.

Summary of the results of transcranial direct current stimulation (tDCS) and transcranial magnetic stimulation (TMS) studies

CitationTarget of the treatmentMeasureBaseline scores mean (SD)