Parkinson’s disease (PD) in elderly people is a frequent neurodegenerative disease. The pathogenesis of PD related to dopaminergic neuron degeneration in the striatum appears to involve genetic and underlying external factors. Oxidative stress, mitochondrial dysfunction, homocysteine, and neuroinflammation have been identified as possible pathways in previous laboratory investigations. Among environmental factors, as it is a potentially modifiable factor, nutrition is one of the most researched fields. The goal of this review is to provide current information about the relationship between diet and risk of PD. From 2000 onwards, we published a systematic overview of the most important research, including prospective studies, nested case-control studies, and meta-analysis. We concentrated on specific nutrients and types of food, alcoholic drinks, uric acid, and dietary trends among dietary variables. These theories have been explored by several studies, but no thorough systematic analysis of the literature has been available to date. For current systematic reviews or meta-analyses of nutrition and PD, the MEDLINE, EMBASE, and WEB OF SCIENCE databases were searched, and one coffee drinking meta-analysis and one antioxidant meta-analysis were found. In addition, microbiota studies have been included since recent results have shown a potential effect on neurodegeneration. In conclusion, there are still many concerns about the association between PD and diet, which may be due to the underlying genetic and gender-specific causes, in addition to methodological variations between studies. However, there is some evidence that uric acid, polyunsaturated fatty acids, coffee, and tea may have a potential protective effect, but mostly in men, while milk products, particularly milk, may increase the risk of PD by means of a contaminant-mediated effect.
The precise contributions to the aetiology of Parkinson’s disease (PD) from genetic and environmental causes remain largely unclear, almost 200 years after James Parkinson’s first recognition of PD (Parkinson, 1817). As the elderly percentage of the world ‘s population rises, concerns about ageing disorders are becoming increasingly important. PD is the second most prevalent neurological condition, and among those over 65 years, its worldwide prevalence is estimated to be around 2 percent (Zhang et al., 2002).
Concordance rates are equally poor in monozygotic and dizygotic twin research, indicating that the aetiology of PD cannot be entirely explained by genetics (Tanner, 2003). In familial PD, unique causal genes have been identified, but these genes in sporadic PD are relatively rare (de Silva et al., 2000; Gwinn‐Hardy, 2002). Therefore, environmental variables are likely to lead to aetiology. In several epidemiological studies, the possible role of diet in the aetiology of PD has been investigated, but the foods and nutrients involved differ from one study to the next. Most have been retrospective case-control (CC) studies; thus, the premorbid diet does not represent exposure calculation. Most were exploratory, evaluating several risk factors for nutrition to produce hypotheses. By direct or indirect effects, particular foods, macronutrients, and micronutrients could contribute to the cause of PD, as with amyotrophic lateral sclerosis-the Parkinsonism dementia complex triggered by neurotoxic cycad seeds (Cox & Sacks, 2002). In other areas of the world, this exposure is rare, but it illustrates food intake as a potential risk factor for PD.
Indirect effects on the body can affect neurodegeneration through diet. For all studies reporting on the relationship between nutritional risk factors and PD, a systematic search of the published literature was performed. No detailed systematic reviews on this subject have been published to date. Evidence is presented for the association of PD with different foods and nutrients. A quantitative or quantitative exposure measure; an effort to determine the information on exposure prior to the onset of the disease.
In particular, the present analysis will investigate the associations between diet and PD in humans. A comprehensive database of animal literature exists.
These were the set protocols for the studies:
Inclusion/ exclusion criteria:
Inclusion: (1) primary researches that thoroughly examine the connection between different foods, nutrients or food groups with respect to PD; (2) articles in English language; (3) outcome of interest revolves around PD; (4) quantitative or semiquantitative exposure measures.
Exclusion: (1) cross-sectional studies lacking CC design; (2) the measure of exposure was subjective (a couple where eating habits differed among each other) (3) the articles that were showing data from different publications, all the data except for the part belonging to the selected time frame, was excluded; (4) studies ecological in nature, were excluded as exposure and disease are not interrelated among same individuals. (5) those statistical analysis where diagnostic criteria were not specified either in medical record or when examined by a physician or neurologist. Therefore, this review will have a specified focus towards the connection of nutrition with PD in man.
EMBASE, WEB OF SCIENCE and MEDLINE databases were the sources for search of meta-analyses or systematic reviews of relationship between nutrition and PD. The articles searched for in these databases were primary research articles and those that contained strong methodology. A great majority of studies showed no significant relationship between nutrition and PD.
To decide whether previous metanalysis or reviews of nutritional risk factors of Parkinson’s disease were on the database, a search was carried out on WEB OF SCIENCE, EMBASE AND MEDLINE from 2000 to 2019. The search was conducted based on search methodology, presentation of results and inclusion criteria. When the review met the criteria of assessment, then it was reported in accordance with additional relevant results and research. For identifying authentic and original contributions to the topic, the search was made based on published sources on the same databases. To identify original contributions to the topic, the same three databases were searched based on published sources (Glasziou et al., 2002; Khan, 2003; Scottish Intercollegiate Guidelines Network, 2015). It was also done by usage of index terms within EMBASE and MEDLINE. Bibliographies that came from primary searches were scanned then additional refences or relevance were taken out. There was no attempt made for publishing unpublished solicit results, neither from key investigator nor conference abstracts. No attempt was made to solicit unpublished results, either directly from key investigators or through conference abstracts.
The studies included were only those that investigated the risk factors underlying for Parkinson’s disease, intervention, the mechanism of action of different nutrients on the brain, and the ones showing adverse effects of malnutrition in patients susceptible to or having Parkinson’s disease. Moreover, these articles were additionally excluded: guidelines, commentaries, animal studies, and nutritional intervention recommendations.
Data collection process
The conduction of subjective quality scoring was not done. Data was extracted from studies by one author (Glasziou et al., 2002). Other guidelines including the Newcastle-Ottawa scale, used for quality assessment of non-randomized studies in metanalyses helped in the identification of study characteristics for the extraction of data (Glasziou et al., 2002; Wells et al., 2000).
Types of participants:
The studies that were included had participants that had these features: patients irrespective of age having been diagnosed with Parkinson’s Disease, under any sort of drug therapy for Parkinson’s Disease and at any duration of PD.
Synthesis of results
Studies that had the following outcome measures were used and reciprocated qualitatively, semi-quantitatively and quantitatively: effects of micro and macro nutrients on brain, neurodegenerative changes in brain due to malnutrition or overnutrition, risk factors of Parkinson’s disease.
Risk of bias
There was no easy assessment technique for the quality of unpublished results, therefore value of such data was limited, however it is probable by the publication bias that negative studies do exist substantially. The studies did not show if the interviewer was blinded throughout the extraction of data to CC status, therefore a chance of interviewer bias remains. In a retrospective CC study, it might not have been possible to avoid a recall bias, however the reporting of results and verification with caution in a cohort is possible.
The review of abstracts and titles of possible meta-analyses and systematic reviews was done including one for antioxidants and coffee consumption (Hernan MA, 2003). Once the exclusion and inclusion criteria were applied, the total count of cohorts was 4 and case-controls was 9. Searches were conducted on MEDLINE, WEB OF SCIENCE AND EMBASE. Articles were found through MEDLINE and EMBASE. No article was found through WEB OF SCIENCE. Other studies were found through the bibliographies of primary research. There is one cross sectional study that is included, as the methods matched the terms of need.
Cohort populations: cohorts from USA were 3 (prospective) and one was added from Netherlands. Two prospective NCC studies were also included, where one was from Leisure world cohort study (Paganini-Hill, 2001). The NHS (Nurses’ Health Study) and the HPFS (Health Professionals’ Follow-up Study) did a recruitment of participants through healthcare occupations via a survey that was mailed with good response but selective population base that restricted the representation of general population. (Table 1)
Table 1: characteristics of cohorts
|Cohort||Authors||Date||Location||Population||Mean age||Range age|
|Framingham study||Fink, J. S., Bains, L. A., Beiser, A., Seshadri, S., & Wolf, P. A. (2001)||Not Available||USA||Massachusetts follow-up||69||–|
|Leisure world cohort||Paganini-Hill (2001)||1981||USA||Retirement community residents||75 ^ 6.1||–|
|Nurses’ health study||Hagan et al. (2016), Ascherio et al. (2001), Chen et al. (2002), Zhang et al. (2002), Chen et al. (2003), Hernan et al. (2003), Chen et al. (2004)||1976||USA||Female registered nurses||30-55|
|Rotterdam study||Willems-Giesbergen, P. C. L. M., De Rijk, M. C., van Swieten, J. C., Hofman, A., & Breteler, M. M. B. (2000)||1990||Netherlands||Community-based recruitment||55 +|
Case-control: heterogenous selection was adapted for methods of CC. Case selection was done from hospitals, patient support groups, neurology clinics, pharmacy databases, and movement disorder clinics. Controls were identified from communities, healthcare databases, spouses, and hospitals. The exclusion was made for dementia commonly. (Table 2)
Table 2: characteristics of case-controls
|Author||Source of cases||Location||Exclusion criteria||Control selection||Exclusion criteria||Matching specification|
|Behari et al. (2001)||Movement disorders in patients consecutively||India||Unreliable information or dementia||Neuro diseases and hospitals||Dementia||Age ^ 3 yrs|
|Benedetti et al. (2000)||PD records||USA||Males||Random / community||Males, tremors||Sex and age ^ 1 yrs|
|Checkoway et al. (2002)||Pharmacy database, diagnostic logs||USA||Medication used that is known to cause Parkinson’s||Enrollees of health cooperative sector||Lack of PD history or neuro disorder history||Matched frequency through clinic, sex, age, GHC enrolment year|
|De rijk et al.||Volunteers from across the country||Netherlands||< 55 age, not living independently, demented||Community||< 55 age, not living independently, demented||None|
|Powers et al. (2003)||Pharmacy database, diagnostic logs, neuro referrals||USA||MMSE < 24||Enrollees of health cooperative sector||MMSE < 24 or neuro disorders||Matched frequency through clinic, sex, age, GHC enrolment year|
|Preux et al. (2000)||In-patients and Out-patients||France||Not lived in the same region||In-patients and Out-patients||Residence||Sex and age ^ 5 yrs|
|Ragonese et al. (2003)||Out-patient at neuro clinics||Italy||MMSE < 24, neuroleptic treatment done within 6 months||Random selection by use of population records.||MMSE < 24 or neuro disorders||Sex and age ^ 2 yrs|
|Tan et al. (2003)||Movement disorder database on random selection||Singapore||Dementia, exclusion of Chinese race||Community health screening program participant||Features of parkinsonian presence||Sex, race and age ^ 0.5 yrs|
|Tsai et al. (2002)||Data bank of hospital movement bank database||Taiwan||Family history, not YOPD, extensive neuro disease||Volunteers in the same hospital||Features of parkinsonian features||Sex and age|
Exposure measurement is qualitative and semi-quantitative or quantitative. The qualitative question required a simple yes and no. or ever v. never. Other measure included relative measuring, e.g. asking about the size of item consumed in comparison to spouse before onset of PD. However, these results are excluded as the quantity is not precisely defined. The semi-quantitative measure is about the usual times food is being taken in a certain time frame with serving size specified. The most precise about frequency of intake and serving sizes are quantitative measures. The estimates for risk were calculated as; highest v lowest tertiles, above v, below median, number of servings, times consumed, quartiles v. quintiles, logistic regression.
The diagnosis of PD was done through clinical signs and symptoms as a conclusive diagnostic test for PD is currently not available. Diagnostic methods in the studies included: interview, medical record interview, physician, and neurological exams.
According to one CC study, there is a non-significant increment in OR with cod liver oil or vitamin intake for PD patients, when taken once a week (Chan et al., 1998). Similarly, another CC study of interest declared no association between PD and vitamin intake (Ghani et al., 2002). Overall, it can be concluded that there are no major relationships between multivitamin usage and PD.
Vitamin A and carotenoids:
Vitamin A and carotenoids have been a topic of interest for their antioxidant properties discussed in animal studies; however, their effects in man have not been under constant demonstration (Institute of Medicine (US) Panel on Micronutrients, 2001; U.S. Department of Agriculture, 2002). The CC study that measured carotenes in total that lie above and below the median showed a report of non-significant positive association. However, results for a-carotene turned out to be non-significant. Two studies were adjusted for sex, age, energy intake and smoking. According to a cohort study (Zhang et al., 2002) there is an inverse relationship between carotenes and PD; however, a CC study include towards a positive association between PD and intake of a-carotenes. Among the four studies of interest, one cohort and two CC studies showed a non-significant association for lycopene (Powers et al., 2003). Similarly, the larger CC reported increased OR (Powers et al., 2003). Lastly, a cohort study showed decrement in RR (Zhang et al., 2002). Further investigation defined three CC studies to be proving positive associations between lutein and PD, however two were also significant (Powers et al., 2003). Moreover, it was seen that the OR showed decrement as the publication date was moved to more recently, and similarly, estimate of non-significance arrives from the largest study. These studies were particularly adjusted for age and sex. Only one cohort showed non-significant decrease in the risk of zeaxanthin and lutein (Zhang et al., 2002). It can be concluded from the data that there are no significant relationships between PD and vitamin A according to CC studies; however, a female cohort and NCC study showed borderline significant increased RR. Similarly, b-, Alpha- and total carotenes have no clear associations with PD. There was a borderline significant increase in OR seen in the case of Lycopene, in accordance with one CC study. CC studies showed association of PD with high intake of Lutein; however, two cohorts denied the association. There was only one study that observed other nutrients related to carotenes, with a significant increase in OR for xanthophylls and xanthin, however the same was not the case with b-cryptoxanthin and cryptoxanthin.
Some research has been conducted to prove possible treatment via vitamin E for PD, however the effects are still questionable (Fariss & Zhang, 2003). There were three ways to assess vitamin E intake: dietary intake, total intake including supplements and intake of foods that contain vitamin E. There are risks reported for quartiles (Powers et al., 2003), quintiles (Zhang et al., 2002), above and below the median, and lastly, a continuous variable mentioned in mg (Boulos et al., 2019). According to a CC study covering food consumption with vitamin E, there is a non-significant increase in risk of PD (Boulos et al., 2019). In a cross-sectional study, there was seen a significant inverse association between vitamin intake per 10 mg/d and PD (Boulos et al., 2019). There was no report of significant risks as shown by the five CC studies conducted for total vitamin E. The trends for sample size or publication were not clear.
In summary, the intake of vitamin E or foods containing vitamin E do not show any significant associations with PD, in accordance with the CC studies. According to the cohort studies, there is a low risk of PD in women and a slight increase in men with high vitamin E intake.
According to the research, except for the NCC, there is no significant association between vitamin C and PD. From the cross-sectional study, there was no association seen between PD and vitamin C per 1000 mg/d. there was a CC study that showed decrement in OR for the higher consumption of food that contained vitamin C (Boulos et al., 2019). There was a reporting for risk in quintiles (Powers et al., 2003; Zhang et al., 2002), tertiles (Paganini-Hill, 2001), median intake, and quartiles (Powers et al., 2003). Among the four CC studies, two showed decreased risks and two showed increase in risk with high intake of vitamin C. There was a high estimate seen from the smallest study, and with increasing sample size, the estimates decreased majorly. These studies were adjusted for sex and age.
The consumption of carbohydrates was reported as: monosaccharides, disaccharides, total carbohydrates, lactose, and by measurement of foods rich in carbohydrates. The reporting of risk estimates was done as quintiles (Chen et al., 2003) and quartiles (Chen et al., 2002). Among the three CC studies measuring total carbohydrate consumption, there was seen a significant increase in OR for PD in one (Chen et al., 2002). However, in accordance with the other two CC studies, there was no significant association and instead there was a reporting of estimates that lie in opposite directions. The HPFS and NHS cohort studies showed an increase in the risk of PD in women and vice versa for men with high total carbohydrate consumption (Chen et al., 2003). Increased adjusted OR were reported by a German CC study for the high quartiles of monosaccharide and disaccharide intake, however only monosaccharide estimates were significant (Sherzai et al., 2016). There was no report for significant risks for cereals or others. One CC study for combined cereal and bread intake showed a non-significant decrease in OR (Sherzai et al., 2016). Measurement of other starch foods along with cereals was done sex-specifically, with no significant associations (Chen et al., 2002). One CC study for individuals that consume French loaf or wheat bread reported a significant decrease in unadjusted OR (Chen et al., 2002). Another CC study among individuals that consume high quartiles of potatoes reported a significant decrease for PD (Sherzai et al., 2016). In conclusion, some CC studies showed high intake of sugars and total carbohydrates is associated with PD. The confirmation of such finding was not given by cohort studies, as they found no effects significantly.
Different studies showed the relationship between PD and subtypes of fat. Intake of animal fat was measure by two CC studies, which found positive significant associations. The PHFS and NHS cohorts found an increase RR in men for PD and decreased RR in women while in comparison of lowest and highest quintiles for animal fat intake (Chen et al., 2003). An HPFS cohort reported non-significant reduction in risk of PD for consumption of vegetable fat in highest quintile (Chen et al., 2003). Similarly, non-significant increase in RR for women was seen in the NHS cohort in the highest quintile (Chen et al., 2003). OR was reported in opposite directions by two CC studies for saturated fat intake. Among the two, the larger one found a significant reduction in OR whereas the smaller one found non-significantly increased OR (Powers et al., 2003). The only examinations of relationships of PD with polyunsaturated, monosaturated and trans-unsaturated fat were done by NHS and HPFS cohorts (Chen et al., 2003). However, no RR was significant. Men have a slight increase risk for monounsaturated fat and women showed decreased risk even for the highest intake. An inverse relationship was reported between PD and arachidonic acid intake (Chen et al., 2003). The RR for EPA, n-3 fatty acids and DHA were under unity, however nonsignificant. A non-significant relationship between linoleic acid and PD was also observed (Chen et al., 2003). In summary, one NCC and two cohort studies did not report any significant associations between fat and types of fats with PD. At the same time, two CC studies showed significant increase in OR of total or animal fat intake for highest levels. There were non-significant risk estimates for fatty acids. No evidence was found to show any relations of cholesterol with PD.
Protein intake measurement was done as total and by-foods’ protein. The risk estimates were reported as The risk estimates for proteins were reported as quintiles (Chen et al., 2003), frequency of consumption (Bretón Lesmes et al., 2014), and quartiles (Chen et al., 2002; Sherzai et al., 2016). The NHS and HPFS cohorts showed RR for protein intake that was non-significant (Chen et al., 2003). In the HPFS cohort, there was reported non-significant increase in RR for men in observation of red and total meat (Chen et al., 2003). However, non-significant decrease in risk was seen in women according to the NHS cohort (Chen et al., 2003). The studies showed significant decrement in PD risk when individuals consumed eggs daily v. never. There was no significant association between chicken consumption and PD (Chen et al., 2003). The same study revealed no association of PD with fish intake solely. In summary, there is no pattern or association between PD and protein intake. The CC study OR were given in both directions, however the cohort RR were near to unity.
The Ca intake was reported in terms of Ca supplementation and total dietary intake. According to a CC studies, there was a non-significant increase in OR while in comparison of lowest and highest quartiles of dietary Ca (Powers et al., 2003). However, overall, there is no significant association between PD and Ca intake.
The intake of iron was measured as: dietary Fe, total Fe (including supplements), and foods with Fe. One study showed increased OR for total Fe consumption (Powers et al., 2003). There is an association possible between PD and Fe. However, further investigation needs to be done.
Fruits, vegetables, and antioxidants:
According to the present hypothesis, high intake of vegetables and fruits has protective effects for PD because of the multi-nutrient supply through them, however there is also a possibility that pesticide usage on them may be of some risk (Barichella et al., 2009). According to the HPFS and NHS cohorts, men have low risk of PD with high vegetable and fruit intake, however the results were non-significant (Chen et al., 2002).
According to three cohorts and one CC, there were no risk estimates significant for lowest and highest quartiles for total dairy intake (Chen et al., 2002; Park et al., 2005). A further specific analysis of dairy products was done in male cohort (Chen et al., 2002). The risk was shown to be elevated in men with respect to consumption of high amounts of dairy products. When milk consumption was adjusted, PD was found to have no apparent association with Ca (Park et al., 2005). In conclusion, the risk estimates for PD in terms of dairy products were not significant in men, however except for consumption of cream cheese once a week, that showed elevated risk.
Our brain is sensitive to alcohol. It acts as a stimulant when in low levels in the blood, however it mainly affects as a depressant. Premature ageing of the brain is associated with alcoholism, that may lead to permanent or temporary destruction of cognitive function (Hagan et al., 2016). There is also a possibility of alcohol contributing to oxidative stress which can affect the brain (Collins, 2002). Despite that, alcohol has also been seen to have protective effects for vascular diseases and neurodegeneration by unknown mechanisms (Gonthier et al., 2004).
Of these CC studies which compared drinkers to non-drinkers, higher proportion was of those that showed inverse association with two being significant as well. All cohorts showed inverse relations. According to the information from five CC studies (Behari et al., 2001; Benedetti et al., 2000; Checkoway et al., 2002; Ragonese et al., 2003; Tsai et al., 2002) and two cohorts (Hernán et al., 2003; Willems-Giesbergen et al., 2000), there was some evidence for the association of PD with alcohol. However, statistical significance was not proven by majority of the studies.
Tea, coffee, and caffeine:
According to the meta-analysis of three CC studies (Benedetti et al., 2000; Paganini-Hill, 2001; Preux et al., 2000) and three cohorts (Ascherio et al., 2001; Fink et al., 2001; Ross, 2000), coffee drinking was seen to be protective for PD. Similarly, the pooled estimates taken for cohort and CC studies, showed no significant difference among each other. However, some of the risk estimates that were protective did not show statistical significance when none were showing risks. Exposure of tea drinking was taken in the form of ‘yes’ and ‘no’ (Preux et al., 2000), cups per day (Checkoway et al., 2002; Paganini-Hill, 2001) and cup-years (Tan et al., 2003). According to the two CC studies there were inverse significant associations with PD for quantitative measures of tea consumption (Checkoway et al., 2002; Tan et al., 2003). All CC and cohort studies reported inverse association between coffee and PD. All the RR and ORs reported using quantitative measures came less or equal to unity. All OR and RR using quantitative measures were less than or equal to unity. Two OR estimates for binary measures of tea consumption were in each direction, however only one suggested positive association significantly with PD. All the CC studies that measured tea consumption in quantitative results showed inverse associations significantly. The evidence, overall, pointed towards an inverse association between caffeine and PD.
According to one CC, there is non-significantly reduced OR for Potassium and vice versa for Sodium and Phosphorus. There is no significant association between Cu, Mn, Mg, Zn or Se and PD when comparing the lowest to highest quartiles (Powers et al., 2003). According to the HPFS male cohort, there is nonsignificant result for supplement intake of more than 10 mg in the same direction. However, the RR was consistent for 0.025 – 9.975 mg that indicated a deficiency of dose response (Chen et al., 2002).
The association of PD and junk food was examined by three studies. According to one CC, there is a non-significant decrease in OR for high junk food intake (Benedetti et al., 2000). Similarly, for the associations between other miscellaneous foods and PD, the OR was mostly non-significant.
Summary of evidence (Table 3):
Table 3: association of nutritive risk factors with PD
|Case-control studies||Cohort studies|
There was no funding provided to support this article.
Conclusion and recommendations
Genetic and environmental factors are likely to be involved in the aetiology of idiopathic PD. Few, if any, mental exposures to PD have been conclusively correlated with the climate. Due to problems with recall and validity, nutrition is a challenging exposure to assess, but because of the potential impacts on PD, it is necessary to continue research in this field.
Retrospective assessment of food intake is problematic. In cohort studies, correlations found in CC studies but not repeated are questionable. To support past findings and to examine new theories, prospective studies are necessary. In the present study, there were no definitive correlations between PD and any of the foods and nutrients. The only two exposures with a reasonably large number of studies in agreement that there could be inverse correlations with PD are coffee and alcohol consumption. However, studies with estimates in the opposite direction and null results are also available.
Previous study findings did not strongly support the main a priori hypotheses regarding nutritional factors and PD. The proposed beneficial effects of antioxidants and, on the other hand, the adverse effects of oxidative stress have not been strongly confirmed by the null results of most fruit and vegetables, vitamins, fats, and Fe research. By consistently documenting a protective association between PD and alcohol consumption, the hypothesis that alcohol could induce oxidative stress was opposed. A priori assumptions concerning the role of certain nutritional factors in PD were not commonly defined. Instead, it included in the study what tended to be the sum or plurality of variables from the questionnaire. Instead of unjustified multiple testing, future research on this subject should include setting out a priori hypotheses about nutrients and their potential role in PD. Several of the studies checked all available food-frequency questionnaire data
This could all be attributable to chance, without hypotheses even though there were any good outcomes. With larger population sizes, false positive findings are less probable, but confirmation in many unrelated studies is more compelling proof that a correlation is true. Before viewing the data, the methods for analysis should also be specified. This will include determining how the risk assessments should be measured.
It is difficult to perform epidemiological research and an optimal analysis is generally unlikely, but the methodology of the study should be as rigorous as possible. The diagnostic criteria listed should be well recognised and secondary Parkinsonism should be omitted. Analytical modifications should be sufficient, and measures taken to select adjustment factors should be reported. The only estimate common in studies was a crude association. For comparisons across studies with further changes depending on the model, an agreed minimum set of changes such as age, sex and smoking will be helpful. Stratification is another choice, but it is difficult to synthesise this into a systematic review.
The conduction of epidemiological studies seems to be difficult and having an ideal study created is near to impossible. However, methodology of the study should be very strong. Similarly, the diagnostic criteria need to be well-recognized and should not include secondary Parkinsonism. When there are adjustments in the analysis, they should be appropriate with careful reporting of steps. A minimal set of adjustments that may include sex, age and smoking can be helpful to create comparisons across studies with dependency on the specific model. the other option that remains is stratification. However, its synthesizing is difficult into a systematic review. Systematic reviews are authentic and valuable for suggesting and evaluating new areas of research and conclude the quality of available information for creation of hypothesis. With strong associations found between PD and alcohol, dairy products and caffeine, biological research is required to further elucidate their roles. The number of studies evaluating nutrients and foods excluding the main micronutrients were less. More evidence can only arrive when more research is pursued in those areas. Future interests may lie on environmental factors that play a part in the contribution of nutrients in PD. The diet-gene interactions have been seen in animals (Mattson, 2003) and is now required to be done in humans. The effects of certain diet factors can be studied to see the outcomes on disease and reveal further interactions. The co-dependency of vascular disorders, nutrients and PD can be a subject of interest as these may or may not contribute to the development of complications in PD (Foltynie et al., 2002). Larger cohorts may tackle the approach used in nutrition and dementia related studies, which also used blood biomarkers in a prospective manner. However, these may necessitate the usage of larger numbers.
Our Healthcare Writing Help team provides high quality writing services in the field done by healthcare professionals: