Creatina para la enfermedad de Parkinson
Resumen
Antecedentes
La enfermedad de Parkinson es uno de los trastornos neurodegenerativos más frecuentes y la disfunción mitocondrial tiene una función importante en la patogénesis. La creatina es eficaz para mejorar la función mitocondrial. Por consiguiente, puede ser útil para retrasar la progresión de la enfermedad de Parkinson.
Objetivos
Evaluar la eficacia y la seguridad de la creatina utilizada sola o como un tratamiento adyuvante para la enfermedad de Parkinson.
Métodos de búsqueda
Se hicieron búsquedas en el registro de ensayos del Grupo Cochrane de Trastornos del Movimiento (Cochrane Movement Disorders Group), CENTRAL (The Cochrane Library 2013, noviembre, número 4), MEDLINE (enero de 1966 hasta el 10 de noviembre de 2013), EMBASE (1974 hasta el 10 de noviembre de 2013) y en dos bases de datos chinas. Se realizaron búsquedas en los registros de ensayos en curso y actas de congresos, se verificaron las listas de referencias y se contactó con los autores de los ensayos incluidos.
Criterios de selección
Ensayos controlados aleatorios (ECA) que compararon creatina versus placebo para la enfermedad de Parkinson.
Obtención y análisis de los datos
Dos autores de la revisión de forma independiente seleccionaron los ensayos para la inclusión, evaluaron su calidad y extrajeron los datos.
Resultados principales
Se incluyeron dos ECA con 194 pacientes. Ambos ensayos compararon la creatina con placebo para la enfermedad de Parkinson y ambos tuvieron limitaciones metodológicas. No hubo pruebas claras de un efecto sobre la función motora (DM ‐0,26; intervalo de confianza [IC] del 95%: ‐4,39 a 3,88; pruebas de baja calidad), las actividades cotidianas (DM 0,37; IC del 95%: ‐1,28 a 2,02; pruebas de baja calidad) ni la calidad de vida después de uno o dos años de tratamiento. Un ensayo informó eventos adversos graves que no se atribuyeron a la creatina. Además, un ensayo observó tasas mayores de efectos gastrointestinales a los dos años de seguimiento.
Conclusiones de los autores
Las pruebas basadas en los efectos de la creatina en la enfermedad de Parkinson están limitadas por el riesgo de sesgo, los tamaños de la muestra pequeños y la corta duración de los ensayos elegibles. Las pruebas no proporcionan una base fiable sobre la cual tomar decisiones de tratamiento. Se necesitan ECA futuros bien diseñados con un tamaño de la muestra más grande y seguimiento a largo plazo para evaluar la creatina para la enfermedad de Parkinson.
PICO
Resumen en términos sencillos
Creatina para la enfermedad de Parkinson
La enfermedad de Parkinson es uno de los trastornos neurodegenerativos más frecuentes y la disfunción mitocondrial tiene una función importante en la patogénesis. La creatina ha mostrado ayudar a mejorar la función mitocondrial y, por lo tanto, puede ser útil para tratar a los pacientes con enfermedad de Parkinson. Los investigadores de la Colaboración Cochrane examinaron las pruebas sobre si la creatina es eficaz y segura para tratar a los pacientes con enfermedad de Parkinson, utilizada sola o como un tratamiento adyuvante, hasta el 10 de noviembre de 2013.
Se incluyeron dos ensayos controlados aleatorios con 194 pacientes que compararon la creatina con placebo en pacientes con enfermedad de Parkinson. El efecto de la creatina sobre la mejora de la función motora, las actividades cotidianas o la calidad de vida después de uno o dos años de tratamiento para la enfermedad de Parkinson fue incierto debido a la baja calidad de los ensayos y al escaso número de participantes reclutados. No se considera que los eventos adversos graves que ocurrieron en los ensayos estén relacionados con la creatina. Sin embargo, un ensayo informó tasas mayores de efectos gastrointestinales (incluidos diarrea, náuseas y estreñimiento) con la creatina.
Authors' conclusions
Summary of findings
| Creatine versus placebo for Parkinson's disease | |||||
| Patient or population: people with Parkinson's disease | |||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of participants | Quality of the evidence | |
| Assumed risk | Corresponding risk | ||||
| Control | Creatine versus placebo | ||||
| UPDRS motor score | The mean UPDRS motor score in the intervention groups was | WMD ‐0.26 (‐4.39 to 3.88) | 194 | ⊕⊕⊝⊝ | |
| UPDRS ADL score | The mean UPDRS ADL score in the intervention groups was | WMD 0.37 (‐1.28 to 2.02) | 194 | ⊕⊕⊝⊝ | |
| *We have provided the basis for the assumed risk (e.g. the median control group risk across trials) in the footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence | |||||
| 1 Both included trials were reported as randomized, double‐blind trials without giving more detail | |||||
Background
Description of the condition
Parkinson's disease is one of the most common neurodegenerative disorders and is characterised by rest tremor, rigidity, bradykinesia and postural imbalance in people with the disease. Deficiency of the neurotransmitter dopamine in the substantia nigra is considered the main reason for the development of these clinical symptoms (Bernheimer 1973). Mitochondrial impairment, oxidative stress and protein mishandling may play an important role in its pathogenesis (Greenamyre 2004). About 1% of the older population (over 60) experience Parkinson's disease worldwide (de Lau 2006; Samii 2004) and it is more common in men than in women (Lees 2009). It has also been estimated that the number of individuals with Parkinson's disease will double in the next 20 years (Dorsey 2007).
Description of the intervention
Levodopa replacement therapy remains the standard treatment for controlling the symptoms of Parkinson's disease. However, the long‐term use of this drug is often associated with the development of motor complications. Up to 40% of individuals with Parkinson's disease may suffer these adverse events after four to six years of levodopa therapy (Ahlskog 2001). Furthermore, the therapeutic effects of levodopa decrease over time and it cannot prevent the progress of the disease (Alves 2008). For these reasons, new treatments have been investigated.
Creatine, a nutritional supplement, has been regarded as an attractive candidate for the treatment of Parkinson's disease since its evaluation by the National Institute of Neurologic Disorders and Stroke (NINDS) Committee (Ravina 2003) and it is widely used for improving exercise performance. It is a guanidine compound which is naturally synthesised in the liver, kidney and pancreas from the amino acids arginine, glycine and methionine (Ryu 2005). It also can be obtained through diet, especially from meat and fish. No safety concerns about long‐term creatine supplementation have been reported (Bender 2008). Studies have demonstrated that it has neuroprotective and antioxidant properties which may be beneficial for neurodegenerative diseases such as amyotrophic lateral sclerosis (ALS) and Parkinson's disease (Lawler 2002).
How the intervention might work
Evidence has shown that impairment of energy metabolism due to mitochondrial dysfunction may be involved in the pathogenesis of Parkinson's disease (Beal 2005; Mattson 1999). Therapeutic strategies which focus on improving mitochondrial function or enhancing energy provision may therefore be beneficial for slowing the progression of Parkinson's disease. Creatine is converted to phosphocreatine in the body and thereafter transfers a phosphoryl group to adenosine diphosphate (ADP) for adenosine triphosphate (ATP) synthesis to increase energy production. In animal models of Parkinson's disease, creatine supplementation is effective in improving mitochondrial function, reducing oxidative stress and protecting against 1‐methyl‐4‐phenyl‐1,2,3,6‐tetrahydropyridine (MPTP)‐induced dopamine depletion in the substantia nigra (Klivenyi 2003; Matthews 1999; Sestili 2006; Tarnopolsky 2001). Meanwhile, results from a randomised futility clinical phase II trial of creatine in early Parkinson's disease did not show it to be inefficacious and suggested that a phase III clinical trial was warranted (NINDS 2006). Another randomised trial (Bender 2006) revealed that, although oral creatine supplementation over two years, at a loading dosage of 20 g daily, followed by 2 g daily for six months and then 4 g daily maintenance, was not associated with significant improvement of overall UPDRS scores, patients treated in the creatine arm tended to need a smaller dose increase of symptomatic dopaminergic therapy. In addition, long‐term creatine supplementation has been shown to be safe and well tolerated (Bender 2008).
Why it is important to do this review
To our knowledge, there was no systematic review of creatine for Parkinson's disease. We performed this systematic review to investigate the efficacy and safety of creatine when used alone or as an adjunctive treatment for Parkinson's disease.
Objectives
To assess the efficacy and safety of creatine used alone or as an adjuvant treatment for Parkinson's disease.
Methods
Criteria for considering studies for this review
Types of studies
Published or unpublished randomised controlled trials (RCTs) comparing monotherapy or adjuvant creatine therapy with placebo were eligible for inclusion. We excluded quasi‐RCTs that allocated patients by, for example, admission number or date of birth. We only included the first period results from randomised cross‐over trials.
Types of participants
Male or female patients of all age groups with a clinical definite diagnosis of idiopathic Parkinson's disease. We accepted the definition established according to the UK Parkinson's Disease Society Brain Bank clinical diagnostic criteria. We excluded patients with secondary parkinsonism, regardless of reasons.
Types of interventions
We assessed the administration of creatine regardless of dosage and duration, either as monotherapy or adjuvant therapy. The control group was placebo. We allowed co‐interventions, such as levodopa or other anti‐parkinsonian medications, if given equally in both groups.
Types of outcome measures
Primary outcomes
Parkinson's disease motor impairments, measured using the Unified Parkinson's Disease Rating Scale (UPDRS) subscale III.
Secondary outcomes
-
Parkinson's disease activities of daily living (ADL), measured using UPDRS subscale II.
-
Parkinson's disease quality of life (QOL), measured using a scale such as the 36‐item Short Form Health Survey (SF‐36) or Parkinson's Disease Questionnaire ‐ 39 (PDQ‐39).
-
Reduction in levodopa dose.
-
Adverse events.
Search methods for identification of studies
Electronic searches
We used the search strategy recommended by the Cochrane Movement Disorders Group to identify relevant articles and searched the following electronic databases, the search strategy for MEDLINE was modified to suit the other databases:
-
The Cochrane Movement Disorders Group Trials Register
-
The Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2013, November Issue 4)
-
MEDLINE (PubMed) (January 1966 to 10 November 2013) (Appendix 1)
-
EMBASE (1974 to 10 November 2013)
-
China Biological Medicine Database (CBM‐disc) (1979 to 10 November 2013)
-
Chinese National Knowledge Infrastructure Database (CNKI) (www.cnki.net, 1979 to 10 November 2013)
Searching other resources
We screened reference lists of included trials and review articles, and contacted authors of included trials for further data. Also we handsearched the abstracts of 12th, 13th, 14th, 15th and 16th International Congress of Parkinson's Disease and Movement Disorders. We contacted Avicena, the pharmaceutical company who produce creatine, to obtain further data and searched the Clinical Trials (www.clinicaltrials.gov), Current Controlled Trials (www.controlled‐trials.com) and Chinese Clinical Trial Register (www.chictr.org) for ongoing trials.
Data collection and analysis
Selection of studies
Two review authors (Xiao Y, Luo M) independently read the titles and abstracts of all studies identified by the search strategy. Once we had retrieved all potentially relevant articles, each review author independently assessed the full text of each article for inclusion. If opinions differed, we resolved disagreements by discussion among all review authors.
Data extraction and management
Two review authors (Xiao Y, Luo M) independently recorded the following information using a data extraction form:
-
Participants: inclusion and exclusion criteria, number of participants in each group, age, gender, baseline comparability, follow‐up, number of withdrawals and reasons for withdrawal.
-
Methods: trial design, randomisation method, allocation concealment, blinding of participants, investigators and outcome assessors.
-
Interventions: administration method, dosage, duration of treatment, co‐intervention, control intervention.
-
Outcomes: primary and secondary outcomes, adverse events.
-
Others: country and setting, publication year, sources of funding, intention‐to‐treat analysis (ITT).
We resolved any disagreements by through discussion between review authors.
Assessment of risk of bias in included studies
Two review authors (Xiao Y, Luo M) independently evaluated the methodological quality of the considered trials. We employed the methods recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). The quality checklist for assessing the risk of bias mainly contains six specific parameters (sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, other bias). Each parameter includes one or more specific items. In all cases, an evaluation of 'low risk' indicates a low risk of bias, an evaluation of 'high risk' indicates high risk of bias, and an evaluation of 'unclear risk' indicates uncertain risk of bias. We resolved any discrepancies by discussion between all review authors. We recorded the final assessment results in 'Risk of bias' tables in Review Manager (RevMan).
Measures of treatment effect
We managed data according to the ITT principle. For dichotomous outcomes, we presented results as risk ratios (RR) with 95% confidence intervals (CI). For continuous data, we presented results as mean differences (MD) with 95% CI.
Unit of analysis issues
For trials with more than two intervention groups, we planned to only include the creatine intervention and placebo group. For cross‐over trials, we planned to only include data from the first period in the analysis.
Dealing with missing data
We planned ITT analysis , further data obtained from authors were analysed as available data.
Assessment of heterogeneity
We evaluated clinical and methodological heterogeneity across the included trials by comparing characteristics of participants, interventions and trial designs.
We evaluated statistical heterogeneity among the included trials using a Chi2 test with α of 0.1, and with the I2 statistic. If the I2 statistic is more than 50%, this indicates substantial statistical heterogeneity (Higgins 2011). If we included a sufficient numbers of trials, we planned to examine the sources of potential clinical and methodological heterogeneity.
Assessment of reporting biases
We planned to assess publication bias with funnel plots.
Data synthesis
We analysed the data using Review Manager (RevMan). Whether we used a fixed‐effect or a random‐effects model mainly depended on the results of the Chi2 test and the I2 test statistic for heterogeneity. Where we found substantial statistical heterogeneity, we adopted a random‐effects model and aimed to examine the sources of heterogeneity. If we did not detect any significant statistical heterogeneity, we used a fixed‐effect model. If clinical heterogeneity was discovered, we planned to perform subgroup analyses.
Subgroup analysis and investigation of heterogeneity
We planned to compare the following subgroups, if we included a sufficient number of RCTs:
-
different dosages of creatine;
-
different durations of treatment (e.g. one year, two years etc).
Sensitivity analysis
We planned to perform the following sensitivity analyses, if we included a sufficient number of RCTs:
-
re‐analysis excluding unpublished trials;
-
re‐analysis excluding trials with a high risk of bias (e.g. without adequate allocation concealment or blinding);
-
re‐analysis using a random‐effects model if a fixed‐effect model was used previously.
Results
Description of studies
See the following sections: Characteristics of included studies; Characteristics of excluded studies; Characteristics of ongoing studies.
Results of the search
We identified a total of 211 references from literature searches up to 10 November 2013. After reviewing the titles and abstracts, we excluded 203 references as they were either not relevant or duplicate publications. Of the remaining eight references, we excluded one trial (Hass 2007) because the trial authors assessed resistance training with creatine for Parkinson's disease without our prespecified outcome measures and one trial (two references; NCT00449865) is still ongoing. We included two trials (Bender 2006; NINDS 2006) in this review (see Figure 1).
Study flow diagram.
Included studies
We included two RCTs reported in five references: Bender 2006 had an efficacy report and a safety report of creatine, while NINDS 2006 had a 12‐month treatment results report, an 18‐month follow‐up safety report and a partially duplicate report of the same trial. We have listed the full details of the included trials in the Characteristics of included studies..
Excluded studies
We excluded Hass 2007 because the trial authors assessed resistance training with creatine for Parkinson's disease without our prespecified outcome measures.
Risk of bias in included studies
In Figure 2 and Figure 3 we have summarized the overall results of all the risk of bias assessments.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included trials.
Risk of bias summary: review authors' judgements about each risk of bias item for each included trial.
Allocation
Trial authors reported both Bender 2006 and NINDS 2006 as randomized, but did not describe the method used to generate allocation sequence and whether the methods of allocation were concealed or not. Therefore both trials were at unclear risk of selection bias.
Blinding
Both included trials were reported as being double‐blind. However, only NINDS 2006 carefully described the method of blinding and was judged adequate. We could not be sure blinding was adequate in Bender 2006.
Incomplete outcome data
In Bender 2006, nine patients (22.5%) in the creatine group and three patients (15.0%) in placebo group dropped out during the trial and the trial authors included all 12 patients in their data analysis. In contrast, NINDS 2006 had only one patient in the creatine group and two patients in the placebo group who dropped out due to lack of efficacy or adverse effects. Bender 2006 was at high risk and NINDS 2006 at low risk of bias.
Selective reporting
We could not confirm whether all prespecified outcomes were reported as we could not obtain protocols of both included trials. Therefore, both were at unclear risk of selective reporting bias.
Other potential sources of bias
Both included trials enrolled a relatively small number of patients: Bender 2006 did not carry out a sample size calculation while NINDS 2006 employed a futility design which would lead to an underestimation of the statistical power in a two sample comparison analysis (Tilley 2006). Furthermore, both trials and the intervention drugs were sponsored by pharmaceutical companies or another organization. Therefore both trials were at unclear risk of other bias.
Effects of interventions
See: Summary of findings for the main comparison Creatine versus placebo for Parkinson's disease
See: summary of findings Table for the main comparison.
Primary outcomes
Improvement of motor function
We assessed improvement of motor function in both included trials (Bender 2006; NINDS 2006) by changes in the UPDRS motor score from baseline to the end of the treatment period. There was no clinically or statistically significant difference between creatine and placebo group in the change in motor function after treatment (MD ‐0.26, 95% CI ‐4.39 to 3.88, two trials, 194 participants; Analysis 1.1), the homogeneity test indicated significant heterogeneity (I2 = 70%), so we applied the random‐effects model.
Secondary outcomes
1. Improvement of ADL
Both Bender 2006 and NINDS 2006 assessed the improvement of ADL after treatment using the change in UPDRS ADL score from baseline. There was no significant difference between both groups in improving ADL during the follow‐up (WMD 0.37, 95% CI ‐1.28 to 2.02, two trials, 194 participants; Analysis 1.2), the homogeneity test indicated significant heterogeneity (I2 = 66%), so we used the random‐effects model.
2. Improvement of QOL
Only one trial (Bender 2006) contributed to this outcome analyses. The trial authors did not find any difference in favour of either creatine or placebo in the change in QOL after treatment as measured using SF‐36.
3. Reduction in levodopa dose
Neither of the included trials reported this outcome.
4. Adverse events
Both included trials reported adverse effects (Bender 2006; NINDS 2006). Unfortunately, we could not combine adverse effects from the two trials because it was difficult to extract precise data on adverse effects from the reports. NINDS 2006 reported an 18‐month safety assessment of creatine, minocycline and placebo. The most commonly reported adverse effects across the treatment groups were nausea (19%), joint pain (25%), and upper respiratory symptoms (bronchitis, coughing, upper respiratory tract infection, sinusitis; 34%). There were 27 serious adverse effects during the 18‐month follow‐up period, seven of which occurred in the creatine group including cardiomyopathy, coronary artery disease, fatal motorcycle accident, myocardial infarction, spinal stenosis surgery and two of chest pain. However, the blinded investigators and independent medical monitor did not consider any to be definitely or probably related to creatine use. Bender 2006 found a higher rate of patients (29%) with gastrointestinal complaints (constipation, diarrhoea, meteorism, nausea) in the creatine group than in the placebo group (6%) at the end of the two year follow‐up. However, this difference was not detected at one year follow‐up. In addition, creatine did not increase the risk of renal function impairment compared with placebo.
Discussion
Summary of main results
Parkinson's disease is one of the most common neurodegenerative disorders. In this systematic review we assessed the efficacy and safety of creatine used alone or as an adjuvant treatment for Parkinson's disease. We included two RCTs (Bender 2006; NINDS 2006) which had a total of 194 patients and compared creatine with placebo for Parkinson's disease. There was no significant difference in favour of either creatine or placebo on improvement of motor function, ADL or QOL after one or two years' treatment for Parkinson's disease although there was statistical heterogeneity in the results of the two RCTs. For the safety outcomes, taking creatine seems to be safe and well‐tolerated: no serious adverse events were reported, except for a higher rate of patients with gastrointestinal complaints in the creatine group compared to the placebo group at two years' follow‐up.
Notably, the two analyses of both included trials were quite heterogeneous. A reasonable explanation may be the clinical heterogeneity: both trials used a different doses of creatine, and the NINDS 2006 trial added minocycline as a co‐intervention.
Overall completeness and applicability of evidence
The current evidence is very limited in terms of size and applicability. The participants were somewhat younger with fewer co‐morbidities than most people with Parkinson's disease in the community, as is often the case in Parkinson's Disease trials.
Quality of the evidence
Both included trials (Bender 2006; NINDS 2006) had some limitations which require consideration. Firstly, neither of the included trials described the method used to generate allocation sequence and whether the method of allocation was concealed or not. Bender 2006 gave no details on how blinding was achieved. Unclear methods of randomisation, allocation concealment and blinding lead to a risk of selection, performance, attrition and detection bias, which may result in an overestimation of intervention effects (Boutron 2004; Schulz 1995). Secondly, both included trials enrolled a relatively small number of patients: Bender 2006 did not carry out a sample size calculation, while NINDS 2006 employed a futility design which would lead to an underestimation of the statistical power in a two sample comparison analysis (Tilley 2006). Thirdly, in Bender 2006, 20% (12/60) of the trial patients discontinued the trial, and the trial authors excluded all of them from the data analysis. Finally, pharmaceutical companies or other organizations sponsored the intervention drugs and both trials, and there may be conflicts of interest. All of the above mentioned limitations may affect the quality of the evidence.
Potential biases in the review process
We undertook an extensive and comprehensive literature search to minimize bias in the review process and only two RCTs met our inclusion criteria. In the preparation of this review, two review authors independently read and screened studies for inclusion, independently completed data extraction and assessed the quality of included trials to minimize potential biases. We attempted to contact the principal investigators of the included trials for data that were insufficiently reported or missing but we are still awaiting responses. We will include any data we receive in an update of this review.
Agreements and disagreements with other studies or reviews
To our knowledge, the efficacy and safety of creatine for Parkinson's disease have not been systematically reviewed previously.
Study flow diagram.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included trials.
Risk of bias summary: review authors' judgements about each risk of bias item for each included trial.
Comparison 1 Creatine versus placebo, Outcome 1 Change in UPDRS motor score from baseline.
Comparison 1 Creatine versus placebo, Outcome 2 Change in UPDRS ADL score from baseline.
| Creatine versus placebo for Parkinson's disease | |||||
| Patient or population: people with Parkinson's disease | |||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of participants | Quality of the evidence | |
| Assumed risk | Corresponding risk | ||||
| Control | Creatine versus placebo | ||||
| UPDRS motor score | The mean UPDRS motor score in the intervention groups was | WMD ‐0.26 (‐4.39 to 3.88) | 194 | ⊕⊕⊝⊝ | |
| UPDRS ADL score | The mean UPDRS ADL score in the intervention groups was | WMD 0.37 (‐1.28 to 2.02) | 194 | ⊕⊕⊝⊝ | |
| *We have provided the basis for the assumed risk (e.g. the median control group risk across trials) in the footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | |||||
| GRADE Working Group grades of evidence | |||||
| 1 Both included trials were reported as randomized, double‐blind trials without giving more detail | |||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Change in UPDRS motor score from baseline Show forest plot | 2 | 194 | Mean Difference (IV, Random, 95% CI) | ‐0.26 [‐4.39, 3.88] |
| 2 Change in UPDRS ADL score from baseline Show forest plot | 2 | 194 | Mean Difference (IV, Random, 95% CI) | 0.37 [‐1.28, 2.02] |
