Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Cost-effectiveness of the add-on exenatide to conventional treatment in patients with Parkinson’s disease when considering the coexisting effects of diabetes mellitus

  • Hsuan-Chih Chen,

    Roles Conceptualization, Data curation, Formal analysis, Methodology, Visualization, Writing – original draft, Writing – review & editing

    Affiliation Graduate Institute of Clinical Pharmacy, College of Medicine, National Taiwan University, Taipei, Taiwan

    ⨯
  • Chen-Yu Wang,

    Roles Conceptualization, Methodology, Writing – original draft, Writing – review & editing

    Affiliations Graduate Institute of Clinical Pharmacy, College of Medicine, National Taiwan University, Taipei, Taiwan, School of Pharmacy, College of Medicine, National Taiwan University, Taipei, Taiwan, Department of Pharmacy, National Taiwan University Hospital Yun-Lin Branch, Douliu, Taiwan, National Center for Geriatrics and Welfare Research, National Health Research Institutes, Yunlin, Taiwan

    ⨯
  • Hsiu-Hsi Chen,

    Roles Conceptualization, Methodology, Writing – original draft, Writing – review & editing

    Affiliation Institute of Epidemiology and Prevention Medicine, College of Public Health, National Taiwan University, Taipei, Taiwan

    ⨯
  • Horng- Huei Liou

    Roles Conceptualization, Methodology, Supervision, Writing – original draft, Writing – review & editing

    valinawang0220@gmail.com

    Affiliations Department of Neurology, National Taiwan University Hospital, Yunlin Branch, Yunlin, Taiwan, Department of Neurology and Pharmacology, College of Medicine, National Taiwan University, Taipei, Taiwan

    ⨯

Abstract

Objective

This study aims to investigate the cost-effectiveness of the add-on exenatide to conventional pharmacotherapy in patients with Parkinson’s disease (PD) when considering the coexistence of diabetes mellitus (DM).

Methods

We used the Keelung and Community-based Integrated Screening databases to understand the medical utilisation in the Hoehn and Yahr stages of patients with PD. A Markov model with 1-year cycle length and 50-year time horizon was used to assess the cost-effectiveness of add-on exenatide to conventional pharmacotherapy compared to conventional pharmacotherapy alone. All costs were adjusted to the value of the new Taiwanese dollar (NT$) as of the year 2020. One-way sensitivity and probability analyses were performed to test the robustness of the results.

Results

From a societal perspective, the add-on exenatide brought an average of 0.39 quality-adjusted life years (QALYs) gained, and a cost increment of NT$104,744 per person in a 50-year horizon compared to conventional pharmacotherapy. The incremental cost-effectiveness ratio (ICER) was NT$268,333 per QALY gained. As the ICER was less than the gross domestic product per capita (NT$839,558), the add-on exenatide was considered to be very cost-effective in the two models, according to the World Health Organization recommendation. Add-on exenatide had a 96.9% probability of being cost-effective in patients with PD, and a 100% probability of being cost-effective in patients with PD and DM.

Conclusion

Add-on exenatide is cost-effective in PD combined with DM. Considering that DM may be a risk factor for neurodegenerative diseases, exenatide provides both clinical benefits and cost-effectiveness when considering both PD and DM.

Citation: Chen H-C, Wang C-Y, Chen H-H, Liou H-H (2022) Cost-effectiveness of the add-on exenatide to conventional treatment in patients with Parkinson’s disease when considering the coexisting effects of diabetes mellitus. PLoS ONE 17(8): e0269006. https://doi.org/10.1371/journal.pone.0269006

Editor: Giuseppe Pignataro, Universita degli Studi di Napoli Federico II, ITALY

Received: December 12, 2021; Accepted: May 13, 2022; Published: August 11, 2022

Copyright: © 2022 Chen et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the article.

Funding: National Taiwan University Hospital Yunlin Branch funded this research (NTUHYL110.I008&X010). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Parkinson’s disease (PD) is a progressive neurodegenerative disease that affects 100–200 per 100,000 people over 40 years of age [1]. Population aging is expected to impose an increasing social and economic burden on society [2]. In the United States, the annual medical cost for a patient with PD ranges from US$12,805 to US$23,101 [3,4]. There are geographic differences. In Asia, the mean annual cost per patient is reported as $3,635, which is lower than that in Australia ($7020), Europe ($3,635), and the United States. Nevertheless, the medical cost for PD is almost one-seventh that of the GDP. The cost of illness due to PD is enormous and increases with the disease progression [5–7].

In addition, the total cost for patients with PD escalates with the progression of the Hoehn and Yahr (H-Y) stage [5]. For a patient, slowing disease progression by 10% would have a net monetary benefit of US$29,001 (US$36,362 including income lost), and progression by 20% would have net monetary benefits of US$60,657 (US$75,981 including lost income) [8]. Approaches that slow the progression of PD may greatly reduce expenditure on society.

Thus, PD management mainly focuses on slowing disease progression and providing pharmacotherapy for symptomatic control. However, no currently available drugs can inhibit disease progression [9]. Exenatide, a glucagon-like peptide-1(GLP-1), is a second-line treatment for type 2 diabetes mellitus (DM) and has demonstrated clinical benefits for patients with PD in randomised controlled trials (RCT) in 2013 [10,11]. When exenatide is added to conventional pharmacotherapy for PD, it delays disease progression [10], but its cost is relatively high.

Furthermore, the relationship between PD and DM has been demonstrated in both epidemiological and molecular biology studies, which show that preceding type 2 DM increases the incidence and progression of PD [12–23]. As a result, the combined use of exenatide with standard PD treatment may have different roles in the population with both PD and type 2 DM compared to the population with only PD. With regard to dual diseases that modify exenatide, cost-effectiveness must be established.

Considering the significant clinical benefits for patients with PD, but the relatively high cost of exenatide, and the relationship between PD and DM, we conducted cost-effectiveness analyses on the add-on exenatide to the conventional treatment in patients with PD when considering the coexisting effect of DM.

Materials and methods

The Markov decision model was used to analyse the cost-effectiveness of add-on exenatide for PD because PD is a lifetime disease and needs chronic pharmacological treatment. Our Markov model was a modified version of those used in previous studies with a cycle length of a year to capture cost-effectiveness [24,25]. The model for assessing cost effectiveness was constructed to represent the real-world situation. The input parameters for the simulation were obtained from the Keelung and community-based integrated screening database (KCIS) [26], the National Health Insurance Research Database (NHIRD) [27], and currently available evidence. The details of PD screening are described elsewhere [28,29]. We obtained information about the characteristics and medical utilisation of patients with PD from the KCIS database and created an appropriate model for subsequent cost-effectiveness analysis [28,29].

Framing the model

Target population.

A cost-effectiveness analysis was conducted with a hypothetical community-based cohort of 1,000 people with health status or DM according to the prevalence.

Study perspective.

In this study, we used the societal perspective cost-effectiveness analysis.

Comparators.

This study includes two interventions. One is conventional pharmacotherapy and the other is conventional pharmacotherapy with add-on exenatide. The former includes conventional therapy for both PD and DM.

The gold standard for treating PD is levodopa and can be combined with dopamine agonists, monoamine oxidase (MAO)-B inhibitors, amantadine, catechol-O-methyl transferase (COMT) inhibitors, and anticholinergic agents, depending on the severity of symptoms. The conventional therapy for DM consists of metformin monotherapy, sulfonylurea monotherapy, combination use of metformin and sulfonylurea, and combination use of metformin and thiazolidinedione (TZD). Metformin is the first-line therapy for patients with type 2 DM. Patients start metformin immediately after they fail to achieve the glycaemic target by lifestyle modification or they may start metformin therapy immediately if they have relatively high and uncontrolled HbA1c levels [30]. As the disease progresses, second-line drugs that could be added to metformin include sulfonylurea, TZD, dipeptidyl peptidase (DPP)-4 inhibitors, sodium-glucose co-transporter-2 (SGLT-2) inhibitors, GLP-1 agonists, and insulin [30].

Time horizon.

A 50-year time horizon was used to assess the cost-effectiveness of the add-on exenatide to conventional pharmacotherapy compared to conventional pharmacotherapy alone.

Discount rate.

The discounting rate was set at 3% with a range of 0%–6%, based on the cost-effectiveness analysis guidelines of the National Institute of Health Technology Assessment (HTA) of Taiwan [31].

Outcome of the model

The outcomes of the models were life expectancy, quality-adjusted life years (QALYs), and costs. By comparing the two interventions, the incremental cost-effectiveness can be calculated for conventional therapy versus add-on exenatide.

Structuring the model

The Markov decision model was used to analyse the cost-effectiveness of add-on exenatide for PD because PD is a lifetime disease and needs chronic pharmacological treatment. The Markov model used by us is a modified version of that used in previous studies with a cycle length of one year to capture cost-effectiveness [24,25].

We assumed that once the patients entered the exenatide group, they continued using exenatide until the end of the study or until death. The symbol at the end of each treatment arm indicates a Markov chain for the process of PD evolving with time, and the two strategies had the same evolving condition.

In the Markov model, data regarding the initial state of the H-Y stage were based on the prevalence in different stages derived from the results from the KCIS databases.

We hypothesised that the progression of PD is irreversible owing to its natural history. This assumption has been adopted in many previous studies [24,25,32–34].

We considered the preceding type 2 DM and PD simultaneously. Because of the distinct natural history of these two diseases, we postulated that DM would precede PD in this model according to the age-specific incidence rate and real-world situations [29,35].

A cost-effectiveness analysis was conducted with a hypothetical community-based cohort of 1,000 people with health status or DM according to their prevalence. The participants entered the cohort at the age of 40 years, and the time horizon of the study was 50 years (i.e. 50 cycles). We defined healthy people as the beginning of the model to represent the progression from health to disease, in which PD incidence is affected by DM, as mentioned earlier.

Model states.

There are thirteen different states in the tree-based Markov model as shown in Fig 1. The alphabet in the nodes behind each state represents the corresponding state of Markov model they would enter. In the model, people started from “normal” state. They would develop to DM, PD or stay in normal in each cycle. H-Y 1 was the initial stage of incident PD patients. Considering the fact that onset age of DM was younger than PD, the comorbidity of DM and PD occurred after DM. The irreversible features of PD in terms of H-Y stage were remained in this model. The background age-specific death rate was considered identical for all state.

thumbnail
Download:
Fig 1. The Markov model for cost-effectiveness analysis for the add-on exenatide effect on PD with DM.

https://doi.org/10.1371/journal.pone.0269006.g001

Likelihood of events.

The probabilities of transitions between states were retrieved from the KCIS database or literature. The upper and lower limits were simply adapted as indicated in the literature. The 95% CI was used as the upper and lower limit for the sensitivity analysis.

Probability of developing disease.

The probabilities of developing PD or DM were retrieved from Taiwanese studies [27,35]. The average and 95% CI were obtained from multiple years of data. The 50-year time horizon of the hypothetical cohort was further divided into 40–49, 50–59, 60–69, 70–79, and > 80 years. The elevated probability of developing PD in preceding DM came from the probability of developing PD multiplied by the age-specific ratio from studies in Taiwan [23,36]. In the age group 40–60, the ratio was 2.20, and 1.5 for above 60 years old.

H-Y stage transition probabilities.

There are four conditions in this section: (1) PD under conventional pharmacotherapy, (2) PD under conventional pharmacotherapy and exenatide, (3) PD comorbid with DM under conventional pharmacotherapy, and (4) PD comorbid with DM under conventional pharmacotherapy and exenatide. Table 1 shows the annual transition rates among the H-Y stages (Fig 2) from which the annual transition probability can be derived using the Markov process model [28]. Note that the slow transition from H-Y 2 to H-Y-3 (λ2) resulted from the slow progression for patients in H-Y 2 stage following the estimated results from Liou et al [28] that patients spent an average of 6.6 years in H-Y 2 stage, which was longer than 2.8 years in H-Y 1 and 1.4 years in H-Y 3.

thumbnail
Download:
Fig 2. Model for estimating H-Y stage transition probabilities.

https://doi.org/10.1371/journal.pone.0269006.g002

thumbnail
Download:
Table 1. H-Y stage transition rate in different condition.

https://doi.org/10.1371/journal.pone.0269006.t001

Probability of death.

(1) Death in PD

The probabilities of death in PD were retrieved from previous studies in Taiwan that used the KCIS and Ilan database [37]. In contrast to other probabilities of death, the deaths in PD were H-Y stage-specific instead of being age-dependent. We did not consider the effects of exenatide on the death of PD due to the lack of long-term trials and to avoid overestimation.

(2) Death in DM

Age-specific probabilities of death in DM were obtained from the study of death in patients with DM in Taiwan with National Health Insurance Research Database (NHIRD) [38]. The 50 years were further divided into 40–49, 50–59, 60–69, 70–74, and > 75 years.

Regarding exenatide, the incidence of DM complications has significantly declined in many randomised controlled trials, but we could not measure exact reduction in mortality because we did not have an exact value for the reduction in complication [39–42]. Consequently, we referred to the projected 10 and 20 life years (LY) gained in the study by Minshall et al. and estimated the reduction in mortality by life table [43,44]. We obtained an estimated efficacy of 0.88 (95% CI 0.86–0.90) in the add-on exenatide group compared to conventional therapy. By directly multiplying age-specific DM death by 0.88, we obtained the probabilities of DM death with add-on exenatide.

(3) Death due to other causes

Age-specific deaths without DM as a cause in Taiwanese people were obtained from Taiwanese government statistics [45]. The time horizon was divided into 40–49, 50–59, 60–69, 70–79, and > 80 years. The death rates of these five timeframes were represented by the probability of death at the ages of 45, 55, 65, 75, and 80 years, respectively. The upper and lower limits were based on the upper and lower ranges of each timeframe.

Cost.

We used the societal perspective cost-effectiveness analysis. Hence, costs, including direct medical costs and indirect costs, were considered. The former includes outpatient clinics, inpatients, examinations, laboratory tests, and drugs. Indirect costs include the loss of productivity. Costs were estimated from the data available from KCIS, Taiwanese literature, and government. All costs were presented in New Taiwanese Dollars (NT$) and adjusted to the value in 2020 according to the medical component of the consumer price index (CPI) in Taiwan [46].

(1) Costs of PD

The medical costs of PD were obtained from the KCIS and categorised according to the H-Y stage. The cost of H-Y 1 in PD comorbid with DM could not be determined due to the relatively small sample size in our cohort; therefore, there were no patients with PD and DM in H-Y 1. We substituted the costs of H-Y 1 in patients with PD only for these missing data because there would be a small difference in cost between early PD and early PD with DM. We also included the costs of home care in H-Y stages 3 and 4+ as part of the direct cost [47]. Regarding indirect costs, we multiplied the average income per person by the productivity loss derived from previous literature. The average income was retrieved from the Survey of Family Income and Expenditure in 2015, and income of 65 years and above was used on account of the characteristics of older age in PD [48].

(2) Cost of DM

We considered the medical costs between 2000 and 2009 in the NHIRD owing to the scarcity of longitudinal trials, especially for comparing the costs of conventional treatment and treatment with exenatide in Taiwan [49]. The costs incurred between 2000 and 2009 were excluded from exenatide because they had not been approved until 2010. We conservatively estimated that the difference between the two strategies was productivity lost by different mortality rates due to insufficient data on the reduction in total medical costs by exenatide. We reckoned age-specific costs by the average costs and reported linear trends in costs of DM in Taiwan, which increased by about NT$15,000 per capita every 10 years [50].

All the input parameters with their plausible ranges and corresponding references are listed in Table 2. The costs of outpatient clinics and hospitalisation, and the total costs of PD and PD combined with DM are presented in Table 3.

thumbnail
Download:
Table 2. Parameter input and data sources for the cost-effectiveness analysis.

https://doi.org/10.1371/journal.pone.0269006.t002

thumbnail
Download:
Table 3. Medical cost of PD only and PD combined with DM per patient during one year.

https://doi.org/10.1371/journal.pone.0269006.t003

Utilities.

All utility values were obtained from the literature. The utilities of different H-Y stages were obtained from a Japanese study, and the utilities of DM and its annual declines were retrieved from the SHIELD longitudinal study in the United Kingdom [56,61]. For patients with both diseases, utilities were calculated by subtracting utilities with DM from those in each H-Y stage based on the assumption that PD and DM were independent of health utilities. The utilities of people stay in normal status were obtained from the Chinese literature [62]. Utility changes by exenatide are estimated by two studies, in which exenatide increases utility by 0.08.

Analysing the model

The base case analysis in this study was conducted from a societal perspective with a cohort of 1,000 patients.

The initial distribution of the patients was based on the prevalence of DM at 40 years of age, and the other was in a normal state. The prevalence of PD was almost zero at the initial age; therefore, we could ignore it. The same distribution was used for both the interventions in this cohort.

The main results were presented as incremental cost-effectiveness ratios (ICERs), including incremental cost per life year (LY) gained and incremental cost per QALY gained.

Unlike in the United Kingdom and the United States, there was no official threshold for the willingness-to-pay (WTP) for a QALY gained in Taiwan. Therefore, the WHO recommendations were applied in our study to check the cost-effectiveness of our results [65]. If the ICER is less than the GDP per capita than the intervention, it is believed to be ‘very cost-effective’, and if the ICER is between 1 and 3 times the GDP per capita than the intervention, it is considered to be ‘cost-effectiveness’. In 2020, Taiwan’s GDP per capita was NT$839,558 [66].

Sensitivity analysis.

One-way sensitivity analyses of all input parameters, including probabilities, costs, proportion of effects of exenatide, utilities, and discounting rate, were performed with the upper and lower limits of each parameter. One-way sensitivity analysis was used to determine whether the uncertainty of the parameters would affect the results of the cost-effectiveness analysis. The results were presented with a tornado plot, where the most influential parameter was lined on the top of the plot, followed by the rest of the parameters, according to the scale of impact on the results.

We performed probabilistic sensitivity analysis (PSA) to test the robustness of the analysis. In the PSA, 1,000 Monte Carlo simulations were performed. The values of each parameter were randomly drawn from the plausible ranges in each simulation. The estimated ICERs were calculated for each simulation, and the simulation results were presented using an ICER plane. A cost-effectiveness acceptability curve (CEAC) was also presented to show the relationship between the WTP threshold and the probability of being cost-effective in each intervention group. Additionally, several scenario sensitivity analyses were conducted to examine the results in special situations.

The situation includes: (1) exenatide reduces PD incidence by 20% in patients with DM, and (2) the effect of exenatide can be applied to the early stages of PD, so we replaced the H-Y transition rates of 1->2 with that of 2->3.

Statistical software.

The analyses were performed using SAS9.4 (SAS Institute, Cary NC), Microsoft ® Excel (Microsoft Corp., Redmond, WA), and TreeAge Pro (TreeAge Software Inc.).

Ethics statement

Our study does not involve individual data. Parameters pertaining to our decision tree were derived from literature. There is no requirement for an IRB approval.

Results

Base-case analysis

From the social perspective (Table 4), the add-on exenatide brought about an average of 0.39 QALYs gained and a cost increment of NT$117,890 per person in a 50-year horizon compared to the conventional pharmacotherapy. The ICER was NT$302,011 per QALY gained. The ICER per QALY gained was less than the GDP per capita of Taiwan in 2020 (NT$839,558), and add-on exenatide was considered to be highly cost-effective according to the WHO recommendation [67]. With the same setting, the add-on exenatide resulted in saving of 0.14 LYs with an ICER per LY saved of NT$818,111.

thumbnail
Download:
Table 4. Results of the cost-effectiveness analyses: Base-case analysis and scenario sensitivity analyses in Aim 3 (PD + DM).

https://doi.org/10.1371/journal.pone.0269006.t004

One-way sensitivity analysis and probabilistic sensitivity analysis

Fig 3 shows a tornado plot of one-way sensitivity for Aim 3. The details of the tornado are described above. The utilities ranked first and second in this analysis because the incidence of PD was relatively small in the population, and the simulative cohort spent most of the time in the normal or DM groups in our model. Therefore, the utility of DM or normal drastically affected the results. However, if GDP per capita is used as the threshold of WTP, none of the values of the input parameters have deviated the results of cost-effectiveness. In the probability sensitivity analyses (Figs 4 and 5), the add-on exenatide had a 100% probability of being very cost-effective with a WTP for GDP per capita.

thumbnail
Download:
Fig 3. Tornado plot of the one-way sensitivity analysis of all input parameters of the cost-effectiveness analysis comparing conventional pharmacotherapy + exenatide with conventional pharmacotherapy in PD with DM.

https://doi.org/10.1371/journal.pone.0269006.g003

thumbnail
Download:
Fig 4. Incremental cost-effectiveness ratio (ICER) scatter plot of the conventional pharmacotherapy + exenatide vs. conventional pharmacotherapy in PD with DM.

https://doi.org/10.1371/journal.pone.0269006.g004

thumbnail
Download:
Fig 5. Cost-effectiveness acceptability curve (CEAC) for the conventional pharmacotherapy + exenatide in PD with DM.

https://doi.org/10.1371/journal.pone.0269006.g005

Scenario sensitivity analyses

The analysis with early-stage effects result in 0.40 QALYs gained and a cost increment of NT$112,904 per person with an ICER of 285,447 per QALY gained, and LYs gained per person with an ICER of 735,379 per LY saved.

Discussion

Our study found that if the government spends an extra cost of NT$117,890 per person for patients with DM in a 50-year horizon, then an average of 0.39 QALYs per person could be gained. The results suggested that add-on exenatide was more cost-effective than conventional pharmacotherapy.

Relationship between DM and PD

The relationship between DM and PD has been studied for several years. There is growing evidence that preceding type 2 DM increases the risk of PD and accelerates its progression [12–16,18–23]. In our model, we hypothesised that individuals may progress to DM first and then to PD. This is confirmed because the age-specific incidences are different between the two diseases in which people are vulnerable to DM since 40 years of age, whereas those above 60 years of age are vulnerable to DM in PD. Moreover, molecular-based studies have indicated that the pathophysiology of DM may contribute to PD through several pathways, in which PI3K/AKT may play a key role [68,69]. Substantial evidence suggests that the loss of AKT signalling is involved in type 2 DM and PD.

Insulin not only regulates glucose homeostasis but also acts as an important homeostatic factor in the brain [70]. It activates its downstream PI3K/AKT pathway, which regulates a variety of important functions that are typically disrupted in PD, including apoptosis, autophagy, inflammation, nerve cell metabolism, protein synthesis, and synaptic plasticity [71]. Studies have shown marked loss of insulin receptor mRNA in the substantia nigra pars compacta (SNpc) of patients with PD and increased insulin resistance compared to age-matched controls [72,73].

AKT acts as a master regulator of cellular function [74]. Results from experimental models indicate that inhibition of AKT signalling leads to dopaminergic cell death, and dysregulation of AKT signalling may affect the expression of alpha-synuclein in PD [75,76].

Although insulin resistance is insufficient evidence that DM increases the risk of PD, it remains to be seen whether brain insulin resistance is due to impaired transit of insulin through the blood-brain barrier or if the neurons themselves become directly insensitive to the actions of insulin, a combination, or both. Taken together, the DM dose has a negative impact on PD.

In our model, the transition probabilities of the H-Y stage are also accelerated by DM, but there is no published literature on how DM severity (such as HbA1c) contributes to PD progression, so we can only consider the condition with or without DM.

Exenatide and conventional pharmacotherapy in DM

As mentioned above, neuroprotection by a GLP-1 agonist comes from the activation of the AKT pathway, which also benefits other insulin-resistant cells in DM. There might be synergic effects of exenatide when treating patients with PD and DM, but we conservatively assumed that ameliorating the progression of PD is the same in PD only and PD comorbid with DM.

In our model, conventional therapy in patients with DM includes metformin, sulfonylurea, thiazolidinedione, and their combination. Several studies have hypothesised that drug use might affect the risk of PD, but so far, there has been an obvious correlation between these drugs and PD. Metformin has been studied for its versatile ability to modify diseases, and its anti-inflammatory and antioxidant properties may be advantageous for PD [68]. However, experimental models have revealed mixed effects on PD. Epidemiological data also suggest neutral outcomes in the risk of PD incidence [22].

A Taiwanese study showed that sulfonylurea might have detrimental effects on the incidence of PD, but the underlying mechanism is uncertain [22]. Moreover, the detrimental effects disappeared in combination with metformin [22]. Some have suggested that the side effects of hypoglycaemia by sulfonylurea are one of the reasons, but more evidence is needed to elucidate the relationship between sulfonylurea and PD.

Taken together, it seems reasonable to assume that the DM drugs used in conventional pharmacotherapy are independent of PD progression. As for the interaction between DM and PD drugs, we also considered their independence due to the lack of published evidence.

Cost of DM

We retrieved the costs directly from the Taiwanese literature, which analyses the total cost of patients with DM with the NHIRD from 2000 to 2009 [49]. The pharmacotherapy of DM included the drugs mentioned above during this period except exenatide which had not been approved until 2010. We assigned the same DM costs in the two strategies because we could not obtain a direct reduction in exenatide intervention. Studies have suggested that DM-related costs are reduced owing to fewer complications [57,63,64]. To avoid model complexity, we did not consider complications, which made the estimation of cost diminishing in DM with exenatide unattainable. We still have cost-effective outcomes even with underestimation; therefore, more cost-effective results could be expected in the future.

Scenario analyses of effects on early-stage and prevention

In scenario analyses, early intervention with exenatide was more cost-effective than base-case analysis. Based on the mechanism of neuroprotection, it is reasonable to assume that PD is prevented. If future work proves that the scenario is true, the cost-effectiveness of exenatide will be enormous.

Overall, we demonstrated the cost-effectiveness of exenatide in a population-based model of PD combined with DM. As growing evidence considers the negative impacts of DM on other neurodegenerative diseases such as AD, treatments that can provide neuroprotection and reverse the deterioration of neuronal cells will be the first choice for those who are genetically or environmentally prone to the development of degenerative nerve diseases [68,77–80]. Exenatide, which can provide clinical benefits and reduce the economic burden, is potentially another multiple disease-modifying drug.

Our study has several strengths. First, several local data are applied to demonstrate cost-effectiveness. Compared to applying data from studies in other countries, the local data were considered to be a more suitable and reliable source for representing the true effect of add-on exenatide in Taiwan. There are many differences in transition probabilities in H-Y stages due to the distinct characteristics of patients and treatment [28,33,53,54,81]. The local data in our model truly reflect the situation of conventional pharmacotherapy treatment in Taiwan. Other input parameters were derived from the NHIRD and studies in Taiwan if applicable. Modifications to the input parameters obtained from Western countries were also made to account for racial disparities. Second, our model is the first to evaluate the cost-effectiveness of the add-on exenatide where DM was considered in the meantime. Such a model design not only reflects the clinical situation in which type 2 DM precedes PD, but also provides future researchers with a model to conduct a cost-effectiveness analysis that considers the effects of exenatide on the prevention of PD. Third, several sensitivity analyses were performed to test the robustness of the results. Our findings supported the view that add-on exenatide is very cost-effective, even though the costs were thought to be relatively small compared to Western countries. For instance, the costs of PD were approximately US$10,146 to 23,101 in the US, whereas the costs in Taiwan were only half to one-fifth of the costs in the US (NT$141052.6) [82]. If the cost of DM reduction by exenatide is considered, the cost-effectiveness will be more prominent.

Our study has some limitations. First, the sample sizes in our PD cohort were small, and we could not obtain patients with H-Y stage 1 who were PD only; therefore, we replaced it with that in PD with DM. In addition, the relatively scarce information on H-Y stage 4+ may provide unrealistic medical utilisation. However, the trends in costs are in line with those in the literature and seem to be reasonable in late-stage PD patients.

The transition rate estimated from our PD cohort may inevitably include patients with DM. Nonetheless, our rates were still the lowest among other studies [53,54,81]. The effects of DM may have been diluted in this cohort. One explanation for this may be that early-stage PD was found by screening, in which accelerated progression brought about by DM was not yet significant.

Second, we did not consider DM development after PD due to an earlier onset age of DM compared to PD. According to these studies, PD alone may affect glycaemia and insulin tolerance. More evidence is needed to confirm the causal relationship between these two diseases. We chose a hypothesis that has been confirmed in many studies that preceding type 2 DM is a risk factor for PD. We did not consider the severity of DM due to the lack of published literature on this topic, and attainable information was insufficient. Our model could be modified if future studies unveiled a more underlying relationship between PD and DM.

Third, we could not directly obtain the costs of DM treated with exenatide because of the lack of local trials in Taiwan. The costs retrieved from the literature in Taiwan may include insulin. However, the results are cost effective. Retrospective or prospective studies are required to evaluate the efficacy and effectiveness of exenatide. Nowadays, a relatively higher price and non-oral type of exenatide have led to undesirability among patients with DM. Patients were started on GLP-1 agonists until the disease progressed to a more severe status in which insulin was needed. Thus, future studies using the NHIRD should be interpreted cautiously.

Fourth, the increase in utilities by exenatide came from a trial in which patients were not well controlled for HbA1c. This might not indicate the utilities that benefit from exenatide when used in the early stage. However, it remained cost-effective when we input the lowest utilities brought by exenatide in the sensitivity analysis.

Fifth, medication adherence was not considered in the model. It is unclear whether the impact of non-compliance would lead to PD and DM progression. In addition, multiple combinations of conventional pharmacotherapy made it difficult to evaluate non-adherence to specific drugs. Our model demonstrated perfect compliance, similar to that in RCT. In addition, insufficient evidence of the age-dependent efficacy of exenatide hinders us from a delicate model.

Lastly, it remains unclear whether the results can be extrapolated to countries other than Taiwan. Nonetheless, this model provides a reference for other countries as well. Finally, our data on cost were based on the NHRID between 2000 and 2009, when the data on PD were collected. This may also affect generalisation.

Add-on exenatide was demonstrated to be very cost-effective in PD combined with DM, which is based on a population-based viewpoint. Considering that DM may be a risk factor for neurodegenerative diseases, exenatide provides both clinical benefits and cost-effectiveness when considering both PD and DM.

References

  1. 1. De Lau LM, Breteler MM. Epidemiology of Parkinson’s disease. The Lancet Neurology. 2006;5(6):525–35. pmid:16713924
  2. 2. Schenkman M, Zhu CW, Cutson TM, Whetten-Goldstein K. Longitudinal evaluation of economic and physical impact of Parkinson’s disease. Parkinsonism & related disorders. 2001;8(1):41–50. pmid:11472879
  3. 3. Kowal SL, Dall TM, Chakrabarti R, Storm MV, Jain A. The current and projected economic burden of Parkinson’s disease in the United States. Movement disorders: official journal of the Movement Disorder Society. 2013;28(3):311–8. Epub 2013/02/26. pmid:23436720.
  4. 4. Huse DM, Schulman K, Orsini L, Castelli-Haley J, Kennedy S, Lenhart G. Burden of illness in Parkinson’s disease. Movement disorders: official journal of the Movement Disorder Society. 2005;20(11):1449–54. Epub 2005/07/12. pmid:16007641.
  5. 5. Spottke AE, Reuter M, Machat O, Bornschein B, von Campenhausen S, Berger K, et al. Cost of illness and its predictors for Parkinson’s disease in Germany. PharmacoEconomics. 2005;23(8):817–36. Epub 2005/08/16. pmid:16097843.
  6. 6. Keränen T, Kaakkola S, Sotaniemi K, Laulumaa V, Haapaniemi T, Jolma T, et al. Economic burden and quality of life impairment increase with severity of PD. Parkinsonism Relat Disord. 2003;9(3):163–8. Epub 2003/02/08. pmid:12573872.
  7. 7. Kaltenboeck A, Johnson SJ, Davis MR, Birnbaum HG, Carroll CA, Tarrants ML, et al. Direct costs and survival of medicare beneficiaries with early and advanced Parkinson’s disease. Parkinsonism Relat Disord. 2012;18(4):321–6. Epub 2011/12/20. pmid:22177623.
  8. 8. Johnson SJ, Diener MD, Kaltenboeck A, Birnbaum HG, Siderowf AD. An economic model of Parkinson’s disease: implications for slowing progression in the United States. Movement disorders: official journal of the Movement Disorder Society. 2013;28(3):319–26. Epub 2013/02/14. pmid:23404374.
  9. 9. Ahlskog JE, Muenter MD. Frequency of levodopa‐related dyskinesias and motor fluctuations as estimated from the cumulative literature. Movement disorders. 2001;16(3):448–58. pmid:11391738
  10. 10. Aviles-Olmos I, Dickson J, Kefalopoulou Z, Djamshidian A, Ell P, Soderlund T, et al. Exenatide and the treatment of patients with Parkinson’s disease. The Journal of clinical investigation. 2013;123(6):2730. pmid:23728174
  11. 11. Aviles-Olmos I, Dickson J, Kefalopoulou Z, Djamshidian A, Kahan J, Ell P, et al. Motor and cognitive advantages persist 12 months after exenatide exposure in Parkinson’s disease. Journal of Parkinson’s disease. 2014;4(3):337–44. pmid:24662192
  12. 12. Cereda E, Barichella M, Cassani E, Caccialanza R, Pezzoli G. Clinical features of Parkinson disease when onset of diabetes came first A case-control study. Neurology. 2012;78(19):1507–11. pmid:22539572
  13. 13. D’Amelio M, Ragonese P, Callari G, Di Benedetto N, Palmeri B, Terruso V, et al. Diabetes preceding Parkinson’s disease onset. A case–control study. Parkinsonism & related disorders. 2009;15(9):660–4. pmid:19356970
  14. 14. Driver JA, Smith A, Buring JE, Gaziano JM, Kurth T, Logroscino G. Prospective cohort study of type 2 diabetes and the risk of Parkinson’s disease. Diabetes care. 2008;31(10):2003–5. pmid:18599528
  15. 15. Hu G, Jousilahti P, Bidel S, Antikainen R, Tuomilehto J. Type 2 diabetes and the risk of Parkinson’s disease. Diabetes care. 2007;30(4):842–7. pmid:17251276
  16. 16. Miyake Y, Tanaka K, Fukushima W, Sasaki S, Kiyohara C, Tsuboi Y, et al. Case–control study of risk of Parkinson’s disease in relation to hypertension, hypercholesterolemia, and diabetes in Japan. Journal of the neurological sciences. 2010;293(1):82–6.
  17. 17. Palacios N, Gao X, McCullough ML, Jacobs EJ, Patel AV, Mayo T, et al. Obesity, diabetes, and risk of Parkinson’s disease. Movement Disorders. 2011;26(12):2253–9. pmid:21739472
  18. 18. Powers KM, Smith-Weller T, Franklin GM, Longstreth W, Swanson PD, Checkoway H. Diabetes, smoking, and other medical conditions in relation to Parkinson’s disease risk. Parkinsonism & related disorders. 2006;12(3):185–9.
  19. 19. Schernhammer E, Hansen J, Rugbjerg K, Wermuth L, Ritz B. Diabetes and the risk of developing Parkinson’s disease in Denmark. Diabetes care. 2011;34(5):1102–8. pmid:21411503
  20. 20. Simon KC, Chen H, Schwarzschild M, Ascherio A. Hypertension, hypercholesterolemia, diabetes, and risk of Parkinson disease. Neurology. 2007;69(17):1688–95. pmid:17761552
  21. 21. Sun Y, Chang Y-H, Chen H-F, Su Y-H, Su H-F, Li C-Y. Risk of Parkinson disease onset in patients with diabetes. Diabetes care. 2012;35(5):1047–9. pmid:22432112
  22. 22. Wahlqvist ML, Lee M-S, Hsu C-C, Chuang S-Y, Lee J-T, Tsai H-N. Metformin-inclusive sulfonylurea therapy reduces the risk of Parkinson’s disease occurring with Type 2 diabetes in a Taiwanese population cohort. Parkinsonism & related disorders. 2012;18(6):753–8.
  23. 23. Xu Q, Park Y, Huang X, Hollenbeck A, Blair A, Schatzkin A, et al. Diabetes and risk of Parkinson’s disease. Diabetes care. 2011;34(4):910–5. pmid:21378214
  24. 24. Johnson SJ, Diener MD, Kaltenboeck A, Birnbaum HG, Siderowf AD. An economic model of Parkinson’s disease: implications for slowing progression in the United States. Movement Disorders. 2013;28(3):319–26. pmid:23404374
  25. 25. Zhao Y, Tan L, Au W, Heng D, Soh I, Li S, et al. Estimating the lifetime economic burden of Parkinson’s disease in Singapore. European journal of neurology. 2013;20(2):368–74. pmid:22978629
  26. 26. Chen TH, Chiu YH, Luh DL, Yen MF, Wu HM, Chen LS, et al. Community-based multiple screening model: design, implementation, and analysis of 42,387 participants. Cancer. 2004;100(8):1734–43. Epub 2004/04/10. pmid:15073864.
  27. 27. Li HY, Jiang YD, Chang CH, Chung CH, Lin BJ, Chuang LM. Mortality trends in patients with diabetes in Taiwan: a nationwide survey in 2000–2009. Journal of the Formosan Medical Association = Taiwan yi zhi. 2012;111(11):645–50. Epub 2012/12/12. pmid:23217600.
  28. 28. Liou HH, Wu CY, Chiu YH, Yen AMF, Chen RC, Chen TF, et al. Mortality of Parkinson’s disease by Hoehn–Yahr stage from community‐based and clinic series [Keelung Community‐based Integrated Screening (KCIS) no. 17)]. Journal of evaluation in clinical practice. 2009;15(4):587–91. pmid:19674213
  29. 29. Chen R, Chang S, Su C, Chen T, Yen M, Wu H, et al. Prevalence, incidence, and mortality of PD A door-to-door survey in Ilan County, Taiwan. Neurology. 2001;57(9):1679–86. pmid:11706111
  30. 30. Association AD. Standards of medical care in diabetes—2016 abridged for primary care providers. Clinical diabetes: a publication of the American Diabetes Association. 2016;34(1):3. pmid:26807004
  31. 31. National Institute of Health Technology Assessment. Guidelines for Medical Technology Assessment Methodology. 2013.
  32. 32. Evans JR, Mason SL, Williams-Gray CH, Foltynie T, Brayne C, Robbins TW, et al. The natural history of treated Parkinson’s disease in an incident, community based cohort. Journal of Neurology, Neurosurgery & Psychiatry. 2011;82(10):1112–8.
  33. 33. Liou HH, Wu CY, Chiu YH, Yen AMF, Chen RC, Chen TF, et al. Natural history and effectiveness of early detection of Parkinson’s disease: results from two community‐based programmes in Taiwan (KCIS no. 11). Journal of evaluation in clinical practice. 2008;14(2):198–202. pmid:18284524
  34. 34. Poewe W. The natural history of Parkinson’s disease. Journal of Neurology. 2006;253:vii2–vii6. pmid:17131223
  35. 35. Chang CH, Shau WY, Jiang YD, Li HY, Chang TJ, Sheu WH, et al. Type 2 diabetes prevalence and incidence among adults in Taiwan during 1999–2004: a national health insurance data set study. Diabet Med. 2010;27(6):636–43. pmid:20546280.
  36. 36. Sun Y, Chang YH, Chen HF, Su YH, Su HF, Li CY. Risk of Parkinson disease onset in patients with diabetes: a 9-year population-based cohort study with age and sex stratifications. Diabetes Care. 2012;35(5):1047–9. pmid:22432112; PubMed Central PMCID: PMC3329814.
  37. 37. Liou HH, Wu CY, Chiu YH, Yen AM, Chen RC, Chen TF, et al. Mortality of Parkinson’s disease by Hoehn-Yahr stage from community-based and clinic series [Keelung Community-based Integrated Screening (KCIS) no. 17)]. J Eval Clin Pract. 2009;15(4):587–91. pmid:19674213.
  38. 38. Tseng C-H. Mortality and causes of death in a national sample of diabetic patients in Taiwan. Diabetes care. 2004;27(7):1605–9. pmid:15220235
  39. 39. Bergenstal R, Li Y, Porter T, Weaver C, Han J. Exenatide once weekly improved glycaemic control, cardiometabolic risk factors and a composite index of an HbA1c< 7%, without weight gain or hypoglycaemia, over 52 weeks. Diabetes, Obesity and Metabolism. 2013;15(3):264–71. pmid:23078638
  40. 40. Klonoff DC, Buse JB, Nielsen LL, Guan X, Bowlus CL, Holcombe JH, et al. Exenatide effects on diabetes, obesity, cardiovascular risk factors and hepatic biomarkers in patients with type 2 diabetes treated for at least 3 years. Current medical research and opinion. 2008;24(1):275–86. pmid:18053320
  41. 41. Ratner R, Maggs D, Nielsen L, Stonehouse A, Poon T, Zhang B, et al. Long‐term effects of exenatide therapy over 82 weeks on glycaemic control and weight in over‐weight metformin‐treated patients with type 2 diabetes mellitus. Diabetes, Obesity and Metabolism. 2006;8(4):419–28. pmid:16776749
  42. 42. Russell-Jones D, Cuddihy RM, Hanefeld M, Kumar A, González JG, Chan M, et al. Efficacy and safety of exenatide once weekly versus metformin, pioglitazone, and sitagliptin used as monotherapy in drug-naïve patients with type 2 diabetes (DURATION-4). Diabetes care. 2012;35(2):252–8. pmid:22210563
  43. 43. Minshall ME, Oglesby AK, Wintle ME, Valentine WJ, Roze S, Palmer AJ. Estimating the long-term cost-effectiveness of exenatide in the United States: an adjunctive treatment for type 2 diabetes mellitus. Value in Health. 2008;11(1):22–33. pmid:18237357
  44. 44. Muennig P, Bounthavong M. Cost-effectiveness analysis in health: A practical approach: John Wiley & Sons; 2016.
  45. 45. Clarke CE, Deane KH. Ropinirole for levodopa‐induced complications in Parkinson’s disease. The Cochrane Library. 2001.
  46. 46. Directorate General of Budget, Accounting and Statistics, Executive Yuan. R.O.C. Consumer Price Commodity Classification Index. 2011.
  47. 47. Workforce Development Agency, Ministry of Labor, Executive Yuan. R.O.C. Foreign Workers to work in Taiwan. 2016.
  48. 48. Directorate General of Budget, Accounting and Statistics, Executive Yuan. R.O.C. 2015 Report on the survey of family, income and expenditure. 2015.
  49. 49. Chang T-J, Jiang Y-D, Chang C-H, Chung C-H, Yu N-C, Chuang L-M. Accountability, utilization and providers for diabetes management in Taiwan, 2000–2009: An analysis of the National Health Insurance database. Journal of the Formosan Medical Association. 2012;111(11):605–16. pmid:23217596
  50. 50. Lin T, Chou P, Tsai S-T, Lee Y-C, Tai T-Y. Predicting factors associated with costs of diabetic patients in Taiwan. Diabetes research and clinical practice. 2004;63(2):119–25. pmid:14739052
  51. 51. Liu W-M, Wu R-M, Lin J-W, Liu Y-C, Chang C-H, Lin C-H. Time trends in the prevalence and incidence of Parkinson’s disease in Taiwan: A nationwide, population-based study. Journal of the Formosan Medical Association. 2016;115(7):531–8. pmid:26123636
  52. 52. Liou HH, Wu CY, Chiu YH, Yen AM, Chen RC, Chen TF, et al. Natural history and effectiveness of early detection of Parkinson’s disease: results from two community-based programmes in Taiwan (KCIS no. 11). J Eval Clin Pract. 2008;14(2):198–202. pmid:18284524.
  53. 53. Smala AM, Spottke EA, Machat O, Siebert U, Meyer D, Köhne‐Volland R, et al. Cabergoline versus levodopa monotherapy: a decision analysis. Movement disorders. 2003;18(8):898–905. pmid:12889079
  54. 54. Davey P, Rajan N, Lees M, Aristides M. Cost-effectiveness of pergolide compared to bromocriptine in the treatment of Parkinson’s disease: a decision-analytic model. Value Health. 2001;4(4):308–15. pmid:11705298.
  55. 55. Cereda M E., PhD M. Barichella MD E. Cassani MD R. Caccialanza MD G. Pezzoli MD. Clinical features of Parkinson disease when onset of diabetes came first. Neurology. 2012;78:1507–11. pmid:22539572
  56. 56. Takuro Shimbo KH, Manabu Takemura and Tsuguya Fukui. Cost-Effectiveness Analysis of Dopamine Agonists in the Treatment of Parkinson’s Disease in Japan. PharmacoEconomics. 2001;18(8):875–86.
  57. 57. Minshall ME, Oglesby AK, Wintle ME, Valentine WJ, Roze S, Palmer AJ. Estimating the long-term cost-effectiveness of exenatide in the United States: an adjunctive treatment for type 2 diabetes mellitus. Value Health. 2008;11(1):22–33. pmid:18237357.
  58. 58. Ratner RE, Maggs D, Nielsen LL, Stonehouse AH, Poon T, Zhang B, et al. Long-term effects of exenatide therapy over 82 weeks on glycaemic control and weight in over-weight metformin-treated patients with type 2 diabetes mellitus. Diabetes Obes Metab. 2006;8(4):419–28. pmid:16776749.
  59. 59. Department of Statistics, Ministry of Interior, Executive Yuan, R.O.C. Abridged life table in Republic of China Area. 2015.
  60. 60. Marešová P, Klímová B, Kuča K. Medical and non-medical costs of Parkinson disease—comparison of Europe, USA, Asia an Australia. Ceska Slov Farm. 2017 Spring;66(1):3–8. pmid:28569512.
  61. 61. Grandy S, Fox KM. Change in health status (EQ-5D) over 5 years among individuals with and without type 2 diabetes mellitus in the SHIELD longitudinal study. Health and quality of life outcomes. 2012;10(1):99. pmid:22917219
  62. 62. Wang H-M, Patrick DL, Edwards TC, Skalicky AM, Zeng H-Y, Gu W-W. Validation of the EQ-5D in a general population sample in urban China. Quality of Life Research. 2012;21(1):155–60. pmid:21505881
  63. 63. Beaudet A, Palmer JL, Timlin L, Wilson B, Bruhn D, Boye KS, et al. Cost-utility of exenatide once weekly compared with insulin glargine in patients with type 2 diabetes in the UK. Journal of medical economics. 2011;14(3):357–66. pmid:21563878
  64. 64. Guillermin A-L, Lloyd A, Best JH, DeYoung MB, Samyshkin Y, Gaebler JA. Long-term cost-consequence analysis of exenatide once weekly vs sitagliptin or pioglitazone for the treatment of type 2 diabetes patients in the United States. Journal of medical economics. 2012;15(4):654–63. pmid:22369345
  65. 65. WH. O. Cost effectiveness and strategic planning (WHO-CHOICE). 2016.
  66. 66. Directorate General of Budget, Accounting and Statistics, Executive Yuan. R.O.C. Common data of national income statistics. 2016.
  67. 67. WH. O. Cost effectiveness and strategic planning (WHO-CHOICE). 2016.
  68. 68. Athauda D, Foltynie T. Insulin resistance and Parkinson’s disease: A new target for disease modification? Progress in neurobiology. 2016;145:98–120. pmid:27713036
  69. 69. Aviles-Olmos I, Limousin P, Lees A, Foltynie T. Parkinson’s disease, insulin resistance and novel agents of neuroprotection. Brain. 2012;136(2):374–84. pmid:22344583
  70. 70. Ghasemi R, Haeri A, Dargahi L, Mohamed Z, Ahmadiani A. Insulin in the brain: sources, localization and functions. Molecular neurobiology. 2013;47(1):145–71. pmid:22956272
  71. 71. Hirsch EC, Jenner P, Przedborski S. Pathogenesis of Parkinson’s disease. Movement Disorders. 2013;28(1):24–30. pmid:22927094
  72. 72. Duarte AI, Moreira PI, Oliveira CR. Insulin in central nervous system: more than just a peripheral hormone. Journal of aging research. 2012;2012. pmid:22500228
  73. 73. Morris JK, Vidoni ED, Perea RD, Rada R, Johnson DK, Lyons K, et al. Insulin resistance and gray matter volume in neurodegenerative disease. Neuroscience. 2014;270:139–47. pmid:24735819
  74. 74. Greene LA, Levy O, Malagelada C. Akt as a victim, villain and potential hero in Parkinson’s disease pathophysiology and treatment. Cellular and molecular neurobiology. 2011;31(7):969–78. pmid:21547489
  75. 75. Canal M, Romaní-Aumedes J, Martín-Flores N, Pérez-Fernández V, Malagelada C. RTP801/REDD1: a stress coping regulator that turns into a troublemaker in neurodegenerative disorders. Frontiers in cellular neuroscience. 2014;8.
  76. 76. Xu Y, Liu C, Chen S, Ye Y, Guo M, Ren Q, et al. Activation of AMPK and inactivation of Akt result in suppression of mTOR-mediated S6K1 and 4E-BP1 pathways leading to neuronal cell death in in vitro models of Parkinson’s disease. Cellular signalling. 2014;26(8):1680–9. pmid:24726895
  77. 77. Spielman LJ, Klegeris A. The role of insulin and incretins in neuroinflammation and neurodegeneration. Immunoendocrinology. 2014;1.
  78. 78. Duarte A, Candeias E, Correia S, Santos R, Carvalho C, Cardoso S, et al. Crosstalk between diabetes and brain: glucagon-like peptide-1 mimetics as a promising therapy against neurodegeneration. Biochimica et Biophysica Acta (BBA)-Molecular Basis of Disease. 2013;1832(4):527–41. pmid:23314196
  79. 79. Hölscher C. The role of GLP-1 in neuronal activity and neurodegeneration. Vitam Horm. 2010;84(13):331–54. pmid:21094907
  80. 80. M Khan N, Ahmad A, K Tiwari R, A Kamal M, Mushtaq G, M Ashraf G. Current challenges to overcome in the management of type 2 diabetes mellitus and associated neurological disorders. CNS & Neurological Disorders-Drug Targets (Formerly Current Drug Targets-CNS & Neurological Disorders). 2014;13(8):1440–57.
  81. 81. Zhao YJ, Wee HL, Chan YH, Seah SH, Au WL, Lau PN, et al. Progression of Parkinson’s disease as evaluated by Hoehn and Yahr stage transition times. Movement Disorders. 2010;25(6):710–6. pmid:20213822
  82. 82. Huse DM, Schulman K, Orsini L, Castelli‐Haley J, Kennedy S, Lenhart G. Burden of illness in Parkinson’s disease. Movement disorders. 2005;20(11):1449–54. pmid:16007641