The role of cognitive reserve in cognitive aging: what we can learn from Parkinson’s disease

  • Nicoletta Ciccarelli
  • Maria Rita Lo Monaco
  • Domenico Fusco
  • Davide Liborio Vetrano
  • Giuseppe Zuccalà
  • Roberto Bernabei
  • Vincenzo Brandi
  • Maria Stella Pisciotta
  • Maria Caterina Silveri
Short Communication


Parkinson’s disease (PD) typically occurs in elderly people and some degree of cognitive impairment is usually present. Cognitive reserve (CR) theory was proposed to explain the discrepancy between the degree of brain pathologies and clinical manifestations. We administered a comprehensive neuropsychological battery to 35 non-demented participants affected by PD. All participants underwent also the Cognitive Reserve Index questionnaire and the Brief Intelligence Test as proxies for CR. Relationships between CR and cognitive performance were investigated by linear regression analyses, adjusting for significant confounding factors. At linear regression analyses, higher CR scores were independently associated with a better performance on Word Fluency (p ≤ 0.04) and Digit Span (backward) (p ≤ 0.02); no associations were observed between CR and other cognitive tests. Our data provide empirical support to the relation between CR and cognitive impairment in PD. In particular, this study suggests that CR may have greater effects on the cognitive areas mostly affected in PD as executive functions.


Cognitive reserve Aging Neuropsychological examination Executive functions 


The concept of cognitive reserve (CR) was proposed to explain the discrepancy between the degree of pathological changes and their clinical manifestations, and has been used to understand individual differences in clinical resilience to brain pathology [1], in particular as regard the ability to optimize and maximize the performance through two mechanisms: recruitment of brain networks, and/or compensation by alternative cognitive strategies [2]. As regards the compensation mechanism, it has been hypothesized that CR can rely on general semantic knowledge, problem solving abilities, and executive functions [3, 4]. Therefore, the effects of CR may be greater on tasks with high executive or semantic demand, as suggested by a recent study including subjects affected by Alzheimer’s Disease or Mild Cognitive Impairment (MCI) [5]. In addition, in normal adults, there are evidences that CR may have a significant influence especially on executive tasks [6]. In particular, the authors found the CR was associated with a better performance in tasks subtended by the dorsolateral prefrontal cortex (as working memory, attention, fluency, and abstract reasoning), suggesting that CR compensation implies the recruitment of these brain circuits.

As regards Parkinson’s disease (PD), empirical evidence suggests the protective effect of education, the most commonly described proxy for CR, on cognitive decline and in reducing risk for developing dementia [7, 8]. However, educational level may not be independent of socioeconomic and gender biases [1]. Moreover, it is not clear if the protective effect of CR on cognitive performance in PD is domain specific [8].

Our aim was to check in PD subjects without dementia if the CR may influence the cognitive performance using a specific and composite tool for measure CR, including not only education but also occupation, lifestyle data, and premorbid intelligence, another key component of CR [9]. Moreover, bearing in mind the model of a relation between CR and the functions subtended by the dorsolateral prefrontal cortex [6], we hypothesized that in a population of PD, whose cognitive impairment is typically dominated by a dysexecutive disorder due to the dysfunction of the cortico-striatal circuits, higher CR may be associated with less severe dysexecutive deficits.



We randomly selected 35 PD participants without dementia (see Table 1 for characteristics) followed at the Center for Medicine of the Aging according to UK Brain Bank criteria. All subjects provided informed consent prior to enrollment.

Table 1

Patient’s demographic characteristics. Cognitive Reserve and Neuropsychological Examination

Demographic and clinical characteristics

M and SD or N (%)


Raw scores (M and SD)

Adjusted scores (M and SD)

N (%) ES ≤ 1a

Age (years)

76.05 (7.11)

RWLT immediate recall

27.03 (8.73)

34.59 (8.42)

19 (54.3%)

Education (years)

9.34 (4.60)

RWLT delayed recall

4.34 (2.59)

6.38 (3.18)

11 (31.4%)


27 (77%)

RWLT recognition (accuracy)

0.90 (0.07)

14 (40%)b

PD history (months)

70.14 (44.93)

Raven’s matrices

23.18 (6.45)

26.22 (6.35)

8/33 (24.2%)

Age at time of diagnosis (years)

71.34 (7.55)

Digit span (forward)

5.11 (0.83)

5.45 (0.92)

6 (17.1%)

Akinetic-rigid vs. tremor-dominant subtype

28 (80%)

Digit span (backward)

3.29 (1.05)

3.76 (1.08)

10 (28.6%)


26.46( 3.11)

Double barrage (accuracy)

0.91 (0.09)

9/32 (28.1%)b


21.82 (9.51)


22.17 (10.54)

26.21 (10.03)

11 (31.4%)


9.54 (7.22)

Rey–Osterrieth figure copy

9.43 (8.46)

13.06 (8.14)

32/34 (94.1%)

Previous depression diagnosis

5 (14%)


CRI score

104.66 (23.60)


TIB score

106.43 (12.47)


ES Equivalent Scores, PD Parkinson’s disease, RWLT Rey’s Words Learning Test, WF word fluency, MMSE Mini Mental State Examination, UPDRS Unified Parkinson’s Disease Rating Scale, GDS Geriatric Depression Scale, CRI Cognitive Reserve Index, TIB Brief Intelligence Test

aES ≤ 1 indicates a performance equal or lower than the fifth centile of the normal population

bES not available; cut-off = 0.90

Exclusion criteria were: early onset of motor symptoms (< 45 years of old), other neurological disorders in addition to PD, an MMSE below the normative cutoff [10], Clinical Dementia Rating (CDR) > 0.5.

At the enrolment, the following demographic and clinical variables were considered for each subject: age, education (years), gender, disease duration (months from diagnosis to enrolment), age at the time of diagnosis, symptoms side at the onset, type of symptom (akinetic-rigid vs. tremor-dominant subtype), previous depression diagnosis, Geriatric Depression Scale (GDS) value, Unified Parkinson’s Disease Rating Scale (UPDRS)-III value (see Table 1).

Neuropsychological and cognitive reserve evaluation

The cognitive profile was obtained by a comprehensive neuropsychological battery (see Table 1). In particular, the following domains were explored: verbal long-term memory by Rey’s Words Learning Test (RAWLT); verbal working memory (digit span forward); reasoning (Raven’s Matrices); and executive functions [word fluency (WF); digit span backward, Rey–Osterrieth figure copy, and double barrage]. In the RAWLT, Recognition Task and Double Barrage accuracy was calculated by taking into account both hits and false alarms [score from 50 (random) to 100 (accuracy at ceiling)] [11].

CR was evaluated by the Cognitive Reserve Index (CRI) questionnaire [2] and the Brief Intelligence Test (TIB) [12]. The CRI questionnaire includes three sections: education, working activity, and leisure time; by CRI, a total score is derived and a result < 85 is classified as low. The TIB test is an Italian version of the National Adult Reading Test (NART) [13]; it provides a good estimate of premorbid intelligence; a TIB score < 93.1 is considered under the normal curve.

Statistical analysis

To explore the association between CRI and TIB (independent variables) and cognitive performance (dependent variables), we performed univariate linear regression analyses for each cognitive test; the analyses were run out with the raw scores obtained at neuropsychological tasks. To verify association between CR and test scoring independently of confounding factors, we also planned multivariate models by adjusting for those demographic and clinical variables showing an association (p values ≤ 0.09) with the tests at univariate analyses. Age and education were not introduced into multivariate analyses, because these variables were taken in account both in CRI and in TIB scoring. Because of collinear relationship between TIB and CRI, these two factors were investigated in separate multivariate models.

A two-tailed p value of < 0.05 was considered statistically significant.


Neuropsychological performance at each cognitive task (both raw and adjusted scores) is shown in Table 1. In particular, nearly, all the participants showed a frank abnormal score at the Rey–Osterrieth figure copy, while within the other cognitive abilities, the performance was more heterogeneous (see the proportion of borderline or pathological Equivalent Scores in Table 1).

As regards CR evaluation (see Table 1), 8/35 (23%) and 6/35 (17%) subjects showed an abnormal TIB and a law CRI score, respectively.

Analyzing the effect of CR on cognitive functions, at univariate analysis only Word Fluency (WF) and Digit span backward showed significant associations. Thus, these two tasks were further investigated in multivariate models. In particular, higher CRI [β = 0.40; 95% confidence interval (CI) 0.01–0.35; p = 0.04] and TIB (β = 0.39; 95% CI 0.03–0.66; p = 0.03) scores independently predicted a better performance at WF, after adjusting for the previous depression diagnosis, and UPDRS-III score; the positive associations for CRI (β = 0.38; 95% CI 0.00–0.03; p = 0.02) and TIB (β = 0.40; 95% CI 0.01–0.06; p = 0.01) were also confirmed for the Digit Span backward, after adjusting for GDS score.

No associations were found between CRI and other cognitive domains, except a nearly significant positive correlation (p = 0.06) with Raven’s Matrices, probably due to an overlapping between CR and fluid and crystallized intelligence.


According to prior evidence [1, 7, 8], our results suggest a role of CR to compensate cognitive decline in PD subjects without dementia. An important strength of our investigation in comparison with the previous studies was to use a composite index for CR. In our knowledge, only a recent longitudinal study used a combined index as proxy for CR, confirming the positive role of educational level and socio-occupational class on cross-sectional cognitive performance; however, as the authors underlined in the discussion, they used only indirect available factors for proxies of recent social engagement, as the amount of telephone use and the number of people that participants know well enough to visit [14]. Noticeably, in our study, we applied a specific and composite tool for measure CR including education, working activity, and leisure time data, in addition to a measure of premorbid intelligence, another key component of CR [9], obtaining consistent results.

Furthermore, as in normal subjects and individuals affected by MCI [5, 6], our data suggest that also in the context of PD, there is a greater influence of CR on executive tasks. Indeed, our sample of patients showed a typical dysexecutive syndrome, but participants with higher CR obtained a significant better performance on executive functions. In particular, we found that higher CR was associated with a better performance only on WF and Digit Span (backward), tasks supposed to be very sensitive to the typical “frontal” damage observed in PD, especially during the early stages of this pathology [15]. Specifically, “frontal” patients tend to be impaired on switching during phonemic fluency tasks [16], and rearranging the order of digits during Digit Span backward [17]. In the clinical setting, these results could help to implement specific and efficient prevention and rehabilitation strategies against cognitive decline in PD.

Notably, our participants also showed episodic memory difficulties, but no association was found between CR and long-term memory tasks. Thus, it is likely that the protective effect of CR in PD cognitive deterioration might be domain specific (that is, on executive abilities).

Moreover, at univariate analyses, also UPDRS-III score and depression showed an association with executive tasks; anyway, at multivariate analyses, only CR confirmed to be associated with cognitive performance; therefore, its effect seems to overshadow the negative role of depression and severity of motor symptoms.

No other clinical variables showed an association with “frontal” functions. In particular, in contrast with the previous studies, we found no associations between cognitive performance and age at symptoms onset [15, 18]. This result could be due to difference in study design and selection criteria, because we excluded subjects with early onset PD obtaining a quite homogeneous sample of subjects as regards age. Similarly, we did not found association between cognitive performance and disease duration: actually, it is very difficult to disentangle the relative effect of age, age at symptoms onset, and disease duration [19], and in one study, exploring this issue only age was significantly associated with incident dementia in PD [18].

We acknowledge that our study has some limitations, as uncontrolled biases can occur in studies based on convenience sample from hospital center rather than a population-based sample [19]. Moreover, our analysis was cross section and did not examine the longitudinal effect of CR; Hindle et al. (2016) observed only a limited effect of CR on dementia risk during a 4 year follow-up. However, other studies showed converse results [8]; thus, further longitudinal surveys are needed to better explore this issue.

In conclusion, our data suggest that higher education in combination with a lifestyle characterized by experiences of mental stimulation may help to cope with PD mental deterioration.


Compliance with ethical standards


No specific funding was received for this study.

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Human and animal rights

The protocol was approved by the local Ethics Committee and all procedures were performed in accordance with the Declaration of Helsinki.

Informed consent

All subjects provided informed consent prior to enrollment.


  1. 1.
    Hindle JV, Martyr A, Clare L (2014) Cognitive reserve in Parkinson’s disease: a systematic review and meta-analysis. Parkinsonism Relat Disord 20:1–7CrossRefPubMedGoogle Scholar
  2. 2.
    Nucci M, Mapelli D, Mondini S (2012) Cognitive Reserve Index questionnaire (CRIq): a new instrument for measuring cognitive reserve. Aging Clin Exp Res 24:218–226PubMedGoogle Scholar
  3. 3.
    Barulli D, Stern Y (2013) Efficiency, capacity, compensation, maintenance, plasticity: emerging concepts in cognitive reserve. Trends Cogn Sci 17:502–509CrossRefPubMedGoogle Scholar
  4. 4.
    Stern Y (2012) Cognitive reserve in ageing and Alzheimer’s disease. Lancet Neurol 11:1006–1012CrossRefPubMedPubMedCentralGoogle Scholar
  5. 5.
    Darby RR, Brickhouse M, Wolk DA et al (2017) Effects of cognitive reserve depend on executive and semantic demands of the task. J Neurol Neurosurg Psychiatry 88:794–802CrossRefPubMedGoogle Scholar
  6. 6.
    Roldán-Tapia L, García J, Cánovas R et al (2012) Cognitive reserve, age, and their relation to attentional and executive functions. Appl Neuropsychol Adult 19:2–8CrossRefPubMedGoogle Scholar
  7. 7.
    Glatt L, Hubble JP, Lyons K et al (1996) Risk factors for dementia in Parkinson’s disease: effect of education. Neuroepidemiology 15:20–25CrossRefPubMedGoogle Scholar
  8. 8.
    Poletti M, Emre M, Bonuccelli U (2011) Mild cognitive impairment and cognitive reserve in Parkinson’s disease. Parkinsonism Relat Disord 17:579–586CrossRefPubMedGoogle Scholar
  9. 9.
    Stern Y (2002) What is cognitive reserve? Theory and research application of the reserve concept. J Int Neuropsychol Soc 8:448–460CrossRefPubMedGoogle Scholar
  10. 10.
    Magni E, Binetti G, Bianchetti A et al (1996) Mini-Mental State Examination: a normative study in Italian elderly population. Eur J Neurol 3:198–202. doi:10.1111/j.1468-1331.1996.tb00423.x CrossRefPubMedGoogle Scholar
  11. 11.
    Grier JB (1971) Non parametric indexes for sensitivity and bias: computing formulas. Psychol Bull 75:424–429CrossRefPubMedGoogle Scholar
  12. 12.
    Colombo L, Brivio C, Benaglio I et al (2000) Alzheimer patients’ ability to read words with irregular stress. Cortex 36:703–714CrossRefPubMedGoogle Scholar
  13. 13.
    Nelson HE, O’Connell A (1978) Dementia: the estimation of premorbid intelligence level using the New Adult Reading Test. Cortex 14:234–244CrossRefPubMedGoogle Scholar
  14. 14.
    Hindle JV, Hurt CS, Burn DJ et al (2016) The effects of cognitive reserve and lifestyle on cognition and dementia in Parkinson’s disease-a longitudinal cohort study. Int J Geriatr Psychiatry 31:13–23CrossRefPubMedGoogle Scholar
  15. 15.
    Broeders M, Velseboer DC, De Bie R et al (2013) Cognitive change in newly-diagnosed patients with Parkinson’s disease: a 5-year follow-up study. J Int Neuropsychol Soc 19:695–708CrossRefPubMedGoogle Scholar
  16. 16.
    Troyer AK, Moscovitch M, Winocur G et al (1998) Clustering and switching on verbal fluency: the effects of focal frontal- and temporal-lobe lesions. Neuropsychologia 36:499–504CrossRefPubMedGoogle Scholar
  17. 17.
    Shimamura AP (2000) The role of the prefrontal cortex in dynamic filtering. Psychobiology 28:207–218Google Scholar
  18. 18.
    Aarsland D, Kvaløy JT, Andersen K et al (2007) The effect of age of onset of PD on risk of dementia. J Neurol 254:38–45CrossRefPubMedGoogle Scholar
  19. 19.
    Aarsland D, Kurz MW (2010) The epidemiology of dementia associated with Parkinson’s disease. Brain Pathol 20:633–639CrossRefPubMedGoogle Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  • Nicoletta Ciccarelli
    • 3
  • Maria Rita Lo Monaco
    • 1
  • Domenico Fusco
    • 1
  • Davide Liborio Vetrano
    • 1
    • 4
  • Giuseppe Zuccalà
    • 1
    • 2
  • Roberto Bernabei
    • 1
    • 2
  • Vincenzo Brandi
    • 1
  • Maria Stella Pisciotta
    • 1
  • Maria Caterina Silveri
    • 3
  1. 1.Center for Medicine of the AgingPoliclinico Gemelli FoundationRomeItaly
  2. 2.Catholic UniversityRomeItaly
  3. 3.Department of PsychologyCatholic UniversityMilanItaly
  4. 4.Karolinska Institute and Stockholm UniversityStockholmSweden

Personalised recommendations