Parkinson’s disease (PD) is a progressive, neurodegenerative disease characterized by both motor and non-mo- tor symptoms. According to the Parkinson’s disease foundation, 7 to 10 million people live with PD worldwide [1] . In America, 60,000 individuals each year are diagnosed with PD, with men being one and a half times more likely to have PD than women [1] . PD is typically associated with four hallmark motor symptoms: resting tremor, rigidity, bradykinesia, and postural instability [2] . In addition to these motor symptoms, several studies have identified the existence of non-motor symptoms and their impacts on persons with PD [3] - [6] . These non- motor symptoms can be divided into five categories, including neuropsychiatric, autonomic, sensory, gastrointestinal, and sleep disorders [7] . Research shows that many of these non-motor symptoms persist from diagnosis through the advanced stages of PD. Unfortunately, these symptoms remain untreated by typical PD medical management and can have a profound impact on quality of life (QOL) [3] -[6] .

One of the most common neuropsychiatric symptoms seen in persons with PD is cognitive impairment (CI). Hely et al. [5] found that 36% of the PD population presented with mild cognitive impairment as opposed to 10.7% - 16.8% in the general population [5] . CI in persons with PD manifests in numerous ways. Impairments in visuospatial function, working memory, long term memory, internal control of attention (the ability to develop encoding and retrieval strategies efficiently), procedural learning, and executive functions are among some of the cognitive deficits commonly identified in persons with PD [8] . Furthermore, the negative effect of CI on QOL in persons with PD has been reported in many studies [3] [4] [9] . In a study investigating the effects of PD on QOL, it was found that CI was associated with lower ratings on the PDQ-39 QOL scale. In this study, CI was found to be a more significant predictor of QOL than even the hallmark features of PD including speech, tremor, rigidity, hypomemia, bradykinesia, gait, and postural instability [3] . Furthermore, Antonini et al. [4] identified CI as one of the PD non-motor symptoms indicative of disease progression and consequently reduced QOL.

In addition to CI, dysfunction of both the cerebrovascular and cardiovascular systems has also often been associated with PD. In persons with PD, evidence of cerebrovascular disease has been found in postmortem brain examination, radiologic neuroimaging, and clinically symptomatic individuals [10] . In postmortem studies, cerebrovascular lesions such as white matter lesions, old ischemic infarcts, lacunes, and cerebral amyloid angiopathy were found in PD brains [11] [12] . Although similar post-mortem findings have been reported in age-matched controls, it is suspected that these vascular changes may have a significant impact on the presentation of non- motor symptoms, such as CI, in persons with PD [10] [11] . In late onset PD, clinically symptomatic vascular pathology in the form of minor stroke (in addition to diabetes and ischemic heart disease) resulted in higher Hoehn & Yahr scores, suggesting that cerebrovascular changes had an effect on PD progression [13] . Finally, several studies have shown that cerebrovascular dysfunction presents in the form of widespread cerebral hypoperfusion in persons with PD [14] - [16] . The presence of these cerebrovascular symptoms and changes in PD has led researchers to consider the potential impact of cerebrovascular disease on persons with PD.

The presence of cardiovascular dysfunction has also been noted in persons with PD. Common cardiovascular symptoms seen in PD include impaired blood pressure regulation, impaired thermoregulation, tachycardia, dyspnea, and peripheral edema [17] . In the PD population, the presence of these cardiovascular symptoms has been associated with higher disability levels [18] . Additionally, studies conducted in the general adult population found cardiovascular risk factors, such as hypertension (HTN) and diabetes [19] as well as cardiovascular symptoms such as orthostatic hypotension (OH), to be associated with cognitive decline [20] .

Considering the relationship found between cardiovascular/cerebrovascular dysfunction and CI in otherwise healthy adults, it is thought that this same relationship may also contribute to the increased incidence of cognitive impairment in the PD population. Thus, the purpose of this systematic review of the literature is to determine the cardiovascular/cerebrovascular effects on cognition in persons with PD. With a relationship between CI and cardiovascular/cerebrovascular dysfunction established, more appropriate medical management can be developed to better treat CI and thus improve QOL in the PD population.

To aid in the assessment and analysis of the seven studies, a table was compiled listing the major components and characteristics of each study (Figure 1). Four out of seven studies included for analysis stated a significant p-value of less than 0.05 [23] [25] [27] [29] . In three of the seven studies, a p-value for significance was not explicitly stated, however, all p-values reported as significant by the authors were less than 0.05 [22] [24] [28] . The Downs and Black Checklist was used to assess quality of evidence in non-randomized studies [21] . The checklist is composed of 27 questions with a maximum total score of 32. Grade of 1 is given to a “Yes”, 0 for “No” or “Unable to determine” for all questions except question five (2 = “Yes”, 1 = “Partially”, 0 = “No”) and question 27, where a possible five points can be awarded [21] . Refer to Table 1 for a listing of Downs and Black total scores for each article reviewed. Table 2 contains a detailed scoring break-down for each article included in this review. The Downs and Black checklist was chosen for its ability to facilitate a more in-depth discussion regarding validity and applicability of selected, non-randomized studies to the research question. After grading discussions, the majority of studies accepted for review and analysis were cross-sectional or retrospective, cross-sectional design.

Risk of bias was assessed using the Downs & Black checklist which probes literature for information regarding internal validity including: accurate statistical analyses, validity of main outcome measures, confounding, randomization, interventional compliance, and blinding of assessors & participants [21] .

Synthesis of study results culminated in identification of five major cognitive domains used to report associations of cerebrovascular and/or cardiovascular effects on cognition in persons with PD: executive function, attention, verbal fluency and memory, visual memory, and working memory.

The primary aim of this review was to determine if there is a clinically significant link between cerebral hypoperfusion, cardiovascular dysfunction, and impaired cognition in persons with PD in order to supplement current diagnostic criteria and symptom management. Disease severity and motor involvement were evaluated in all seven studies using the Unified Parkinson’s Disease Rating Scale (UPDRS) and its subscales [22] - [28] . Additionally, five of the seven studies used the Hoehn & Yahr scale to further measure motor symptom severity [22] [23] [25] [27] [28] . A battery of neuropsychological tests were used to assess different dimensions of cognition, the most common being the MMSE [22] [23] [25] - [28] . The MMSE is a validated test for global cognition in Parkinson’s disease [50] . Dubois et al. [50] suggest that while the MMSE is an efficient exam for global cognition, it does not appropriately assess executive function deficits. Age and education level also impact MMSE scores, making it a less reliable tool for the PD population [50] . For these reasons, we eliminated three studies from our review that used MMSE as the primary measure of cognitive function.

This systematic review did not establish a causative correlation between cognitive impairment, brain perfusion, or cardiovascular dysfunction in persons with PD. However, it established that a relationship exists between these variables, the extent of which requires further research. Parkinson’s disease is a progressive, degenerative disorder of the central nervous system, therefore, it cannot be determined with certainty whether cognitive impairment is due to altered blood flow to the brain or the ongoing disease process [51] [52] . Nonetheless, OH has been associated with cognitive decline in dementia, PD, and PDD [52] .

OH is common in persons with PD, with up to 20% of patients experiencing symptoms [53] . This finding could be attributed to disease progression, dopaminergic drugs, or a failure of the autonomic system [52] . These factors differentiate the presentation of OH in PD from the presentation of OH caused by the ageing process [53] . As orthostasis is a common adverse effect of dopaminergic drugs [10] , several of the studies in this review attempted to control for these effects by converting to equivalent doses of standard levodopa [22] [24] [28] , restricting medication for 12 hours prior to evaluation [22] [23] , examining during “off times” [27] , or using a drug-naive population [25] [26] . Overall, the results of the literature reviewed indicate that cognitive test scores were lower in those persons with PD who met the criteria of OH and/or impaired perfusion. Three studies found a significant correlation between decreased attentional scores and comorbid OH in persons with PD [22] [27] [28] . That being said, diminished attention scores in persons with PD may be indicative of greater cognitive impairment.

Cerebral hypoperfusion is not the sole perfusion dysfunction impacting cognition in persons with PD. In a study by Kim et al. [25] , persons with PD demonstrating supine hypertension (SH), defined as systolic BP ≥ 150 mmHg or diastolic BP ≥ 90 mmHg in supine, were found to have decreased cognitive function, specifically impaired visuospatial memory. This was overwhelmingly true of those patients who presented with both SH and OH. None of these persons had normal cognition, and they demonstrated the most severe cognitive impairments when compared to the entire study sample [25] . This finding shows that altered cerebral perfusion secondary to HTN or OH is correlated with decreased cognitive function. Jones et al. [24] found that HTN has an influence on executive function and memory recall that goes beyond the deficits produced by the typical PD disease process. They hypothesized that while these deficits are small, they may become more conspicuous and profound as the disease progresses [24] .

Verbal memory has been shown to be involved with the larger, more encompassing memory network, including areas of the brain such as the precuneus, inferior parietal lobe, and superior temporal gyrus [26] . Similar to persons with Alzheimer’s dementia, persons with PD demonstrated that a lack of perfusion in the posterior parietal limbic areas and the association areas of the frontal lobe have a negative effect on cognition scores [26] . SPECT scans of persons with PD support this finding, indicating that hypoperfusion affects posterior areas of the brain and results in similar verbal deficits [26] .

The same posterior hypoperfusion pattern was found by Derejko et al. [51] . The most identifying feature on SPECT scans of persons with PDD was posterior hypoperfusion, a pattern typically documented in Alzheimer’s disease [51] . Additionally, the authors observed marked hyperperfusion of the left thalamus in persons with PDD [51] . Thus, disease-related cognitive decline may not only be due to structural changes in the brain, but also deficits in cardiovascular or cerebrovascular systems [51] . These findings indicate that brain perfusion mapping has the potential to be an indispensable tool for predicting and tracking declining cognitive function in persons with PD [51] .

This review had several limitations, the first of which was the lack of standardization for on/off times for dopaminergic medications. As previously stated, all studies attempted to control for pharmacological effects, but the different methods used may have confounded the results. Levodopa (L-dopa) can decrease motor symptoms due to increased synthetic dopamine in the system during “on” times [54] . These motor improvements could affect the disease severity measurements, especially the motor subscale of the UPDRS. Correlations of disease severity and cognitive decline could be confounded if patients were evaluated during L-dopa “on” times versus “off” times. More importantly, BP may be elevated when a patient is in the “off” state and BP may drop to symptomatic levels when the patient is in the “on” state [54] . L-dopa has a known hypotensive effect [55] , so if persons with PD were evaluated during “on” times, hypotension might have been a pharmacologic side effect rather than a symptom of cardiovascular dysfunction. This hypotensive effect has been shown to be protective for those with a history of HTN [10] , but for the purposes of this review, it may have confounded the results. Blunting HTN with L-dopa could also change the results of studies that found a relationship between HTN and cognition, even if all other HTN medications were controlled [24] . In the future, a standard for testing during on/off times for L-dopa would strengthen results.

The most significant limitation of this review is the inability to distinguish the origin of cognitive decline in persons with PD. Cognitive decline appears to be correlated with later onset and greater severity of motor symptoms [22] [23] . This correlation could be a result of disease progression rather than cardiovascular dysfunction.

For example, widespread Lewy body pathology has been known to cause fluctuating cognition, dementia, and damage to the anterior cingulate cortex (ACC) [22] . In a study by Critchley et al. [56] , a direct link between ACC activity and sympathetic output was found. Thus, autonomic dysfunction and altered cognition in persons with PD may perhaps be a result of underlying Lewy body disease (LBD) and not cardiovascular risk factors [57] .

A final limitation is the use of drug-naive persons with PD to study cognition [25] [26] . As previously stated, anti-Parkinsonian drugs could affect HTN and consequently cognition [10] , therefore, these studies can definitively say their results are not due to adverse effects of L-dopa. However, that advantage may be nullified by the possibility of an inappropriate diagnosis of PD. Quinn [58] described the differential diagnoses for idiopathic PD and the unfortunate frequency at which PD is misdiagnosed, even by specialists. Without testing for the response to L-dopa, these drug naive populations could potentially be showing signs of LBD or other Parkinsonian syndromes. Kim et al. [25] stated that their sampled population may fall on the continuum between LBD and PD, thus, these results may not be attributable to PD alone.

Having established that impaired perfusion negatively impacts cognitive function, it can be inferred that those persons with PD who also have altered BP will display cognitive impairments. It is the contention of this literature review that persons with PD and co-morbid cerebrovascular or cardiovascular impairments will benefit from early cognitive screening with an appropriate neuropsychological testing battery. Early testing would allow for more appropriate interventions with the intention of slowing or monitoring cognitive decline.

With a relationship between cerebrovascular and cardiovascular dysfunction and cognition established, appropriate management of cognition in persons with PD can be more thoroughly considered. Knowing that this relationship exists, the identification of cardiovascular risk factors, such as OH and HTN, can be utilized to warrant further neuropsychological testing. Currently, in persons with PD the MMSE is utilized as the primary measure for neuropsychological screening [50] . However, recent literature suggests that this test is not accurately able to identify cognitive impairment in persons with PD [59] . In a study by Burdick et al. [59] , it was found that more than half of persons with PD who scored above normal on the MMSE, also scored below normal on the Montreal Cognitive Assessment. Sixty-seven percent of those with normal MMSE scored 1.5 standard deviations below normal in at least 1 other cognitive test. Sixty-four percent of persons with PD deemed cognitively normal by the MMSE scored 1.0 SD below normal on at least 2 other cognitive measures. Furthermore, this study concluded that if the MMSE were to be the only test utilized to detect dementia in persons with PD, then as many as 55% of dementia cases would have gone undetected in this population [59] . Thus, considering the prevalence of cognitive impairment and its impact on quality of life in persons with PD, it is imperative that appropriate neuropsychological testing be utilized to aide in early detection of cognitive impairment, especially in cases of individuals with previously identified risk factors.

With risk factors identified and appropriate neuropsychological testing instituted, further management of cognitive impairment in persons with PD can be explored. In a study by Hohler et al. [60] , conservative management of OH in the form of anti-hypertensive medications, maintaining adequate hydration, increasing salt intake, and wearing compression stockings resulted in improvements in the cognitive component of the Functional Independence Measure. In addition to these conservative techniques, other studies have researched the effects of physical activity in treating cognitive impairment in persons with PD. In a study by Ridgel et al. [61] , the effects of passive leg cycling on cognition were studied over the course of four exercise sessions. Following bouts of passive leg cycling, no improvements were noted in visual search or motor speeds, however, passive cycling was found to improve set-shifting, a component of executive functioning (as identified by the Trail Making Test A and B) [61] .

In another study by Tanaka et al. [62] the effects of a multimodal exercise session on executive function was studied in the PD population. Significant improvements were found in executive function in the training group over the control group (as identified by the Wisconsin Card Sorting Test [63] ). Specifically, improvements in abstraction and mental flexibility were noted [62] . Both of these studies hypothesize that physical activity may be able to positively impact the circulatory system in such a way that executive function is consequently improved [61] [62] . Ridgel et al. [61] , suggest that the physiologic effects typically associated with exercise (increased cardiac output, stroke volume, heart rate, decreased vagal tone) may function to treat the cardiovascular risk factors identified to affect cognition in person with PD. Likewise, Tanaka et al. [62] hypothesized that exercise in persons with PD may result in increased blood flow to the brain, thus, treating the cerebrovascular perfusion deficits associated with altered cognition in this population. Further research, including detailed imaging in the form of perfusion scans during physical activity, may be warranted in order to better support the aforementioned hypothesis that cognitive impairment in PD can be directly treated by the altered circulation and perfusion patterns identified following exercise. With this information determined, the inclusion of health care professionals such as physical therapists will become instrumental in the treatment of cognitive impairment in persons with PD.

Scopus, PubMed, CINAHL, and citations from selected articles

Scopus

Search terms: autonomic [Ti, Ab, keyword], cognitive [TI], cardiovascular [TI, AB, keywords] (54)

autonomic [Ti, Ab, keyword], cognitive [TI], Parkinson [TI, AB, keywords] (29)

PubMed

Search terms: ((cerebrovascular) and Parkinson’s Disease) and Cognitive (175)

cerebrovascular* [TI/AB] or hypoperfusion* [TI, AB] and Parkinson’s [TI, AB] and cognition* (body) (39)

cognition and hypoferfusion and autonomic (17)

“blood pressure” * (TI/AB) and cognitive* (TI/AB) and Parkinson’s (TI/AB) (18)

Cardiovascular cognitive parkinsons (65)

CINAHL

Search Terms: autonomic* [TI, AB] and cognitive*[TI, AB] and Parkinson’s [keyword] (15)

autonomic* [TI, AB] and cognitive* [TI, AB] (10)

autonomic [TI] and Parkinson’s [TI] (10)

From Idiaquez, J., Benarroch, E.E., Rosales, H., Milla, P. and Rios, L. (2007) Autonomic and Cognitive Dysfunction in Parkinson’s Disease. Clinical Autonomic Research, 17, 93-98. http://dx.doi.org/10.1007/s10286-007-0410-7

Goldstein, M.A. and Silverman, M.E. (2006) Autonomics and Cognition. Clinical Autonomic Research, 16, 86- 89.

From Ridgel, A.L., Kim, C.H., Fickes, E.J., Muller, M.D. and Alberts, J.L. (2011) Changes in Executive Function after Acute Bouts of Passive Cycling in Parkinson’s Disease. Journal of Aging and Physical Activity, 19, 87-98.

Tanaka, K., Quadros Jr., A.C., Santos, R.F., Stella, F., Gobbi, L.T. and Gobbi, S. (2009) Benefits of Physical Exercise on Executive Functions in Older People with Parkinson’s Disease. Brain and Cognition, 69, 435-441. http://dx.doi.org/10.1016/j.bandc.2008.09.008

From “Related citations in PubMed” when searching Slobodin, T.N. (2012) Importance of Vascular Factors for Cognitive Rehabilitation in Parkinson’s Disease Patients. Likars’ ka sprava/Ministerstvo okhorony zdorov’ia Ukrainy, 7, 176-179.

Szirmai, I. and Kovacs, T. (2002) Parkinson Syndrome and Cognitive Disorders. Deggyogyaszati Szemle, 55, 220-225.

Common abbreviations

PD: Parkinson’s Disease;

QOL: Quality of life;

CI: Cognitive impairment;

SPECT: Single photon emission computed tomography;

OH: Orthostatic hypotension;

HTN: Hypertension;

PDD: Parkinson’s disease dementia;

MMSE: Mini-mental state exam;

BP: Blood pressure;

SH: Supine hypertension;

UPDRS: Unified Parkinson’s disease rating scale;

L-dopa: Levodopa;

LBD: Lewy body disease.