Downloaded from http://jnnp.bmj.com/ on February 8, 2015 - Published by group.bmj.
com
Research paper
Poor attentional function predicts cognitive decline in
patients with non-demented Parkinsons disease
independent of motor phenotype
J-P Taylor, E N Rowan, D Lett, J T OBrien, I G McKeith, D J Burn
Institute for Ageing and Health,
Newcastle upon Tyne, UK
Correspondence to:
Dr J-P Taylor, Institute for
Ageing and Health, Wolfson
Research Centre, Newcastle
General Hospital, Westgate
Road, Newcastle upon Tyne NE4
6BE, UK;
john-paul.taylor@ncl.ac.uk
Received 22 February 2008
Revised 29 April 2008
Accepted 2 June 2008
Published Online First
27 June 2008
1318
ABSTRACT
Background: Postural instability gait difficulty (PIGD)
motor phenotype in Parkinsons disease (PD) is associated
with a faster rate of cognitive decline than in tremor
dominant cases and may be a risk factor for incident
dementia. People with PD display attentional deficits;
however, it is not clear whether attentional deficits in
patients with non-demented PD are associated with (i)
PIGD phenotype and/or (ii) subsequent cognitive decline.
Aims: (i) To examine rates of cognitive decline (Mini-
Mental State Examination (MMSE) and Cambridge
Cognitive Examination (CAMCOG)) over 3 years in
subjects with non-demented PD aged over 65 years, (ii)
using Cognitive Drug Research computerised assessment
test battery, determine the rate of change in power of
attention (PoA) scores over time (sum of mean choice
reaction time, simple reaction time and digit vigilance
reaction time scores), (iii) determine whether the PIGD
phenotype is associated with changes in PoA and (iv)
establish whether baseline PoA is associated with
subsequent cognitive decline.
Results: 14 subjects (38%) were classified as PIGD
motor phenotype at baseline. Cognitive decline was
greater in PIGD compared with non-PIGD subjects
(p(0.02). PIGD phenotype was not associated with
baseline PoA score although it was associated with
subsequent worsening in PoA (mean PoA increase/year:
non-PIGD subjects 11.4 ms; PIGD subjects 244.0 ms;
p = 0.01). Higher baseline PoA scores were associated
with greater cognitive decline (MMSE, p = 0.03;
CAMCOG, p = 0.05) independent of PIGD status.
Conclusion: PIGD phenotype and attention deficits are
independently associated with a greater rate of cognitive
decline in patients with non-demented PD. We propose
that subtle attentional deficits in patients with non-
demented PD predict subsequent cognitive impairment.
The occurrence of dementia in Parkinsons disease
(PDD) increases mortality and has a significant
adverse impact on function and quality of life.13
The cumulative incidence of PDD may be as high
as 80%.4 A multiplicity of risk factors have been
proposed for the emergence of PDD5; consistent
predictors include older age, severity of motor
symptoms, executive dysfunction and poor verbal
fluency.4 68 In addition, motor phenotype, speci-
fically the non-tremor dominant phenotype that
includes postural gait instability difficulty
(PIGD), is associated with a faster rate of
cognitive decline in PD and higher incident risk
of dementia.9 10 It has been postulated that the
PIGD motor phenotype arises as a result of
additional dysfunction in non-dopaminergic
J Neurol Neurosurg Psychiatry 2008;79:13181323. doi:10.1136/jnnp.2008.147629
systems, in particular neurodegeneration within
the cholinergic system.911
Even before the development of dementia,
cognitive deficits occur in patients with non-
demented PD and encompass a wide range of
domains whose neural substrates may include
fronto-subcortical and temporo-parietal circuits.8 12
Diverse neurotransmitter systems have been impli-
cated, but cognitive sequelae in PD and incident
dementia are particularly dependent on non-
dopaminergic systems, notably the cholinergic
system.5
Attentional deficits are an important component
of cognitive impairment in PD.13 Furthermore,
attentional impairment in PDD is a strong
predictor of the ability of an individual to perform
their activities of daily living.14 It is unclear
however whether or not attentional deficits in
PD identify individuals who will develop dementia;
this postulate has biological plausibility given the
fact that the same neural substrates (eg, fronto-
subcortical and non-dopaminergic systems)
involved in the mediation of attention are known
to be compromised in PD and the associated
dementia.5 15
Postural control and gait is dependent on the
integrity of attentional systems, and interplay
between these systems may be compromised in
disease states such as PD.16 Gait function is
adversely affected in patients with PD tested using
dual task paradigms which impose high attentional
demands.17 It is unclear whether patients with PD
with a PIGD motor phenotype are more predis-
posed to attentional deficits.
In this study, we followed-up a cohort of
patients with non-demented PD over 3 years.
The aims were: (i) to examine rates of cognitive
decline (Mini-Mental State Examination (MMSE)
and Cambridge Cognitive Examination
(CAMCOG)), (ii) using the Cognitive Drug
Research (CDR) computerised assessment test
battery to determine the rate of change in power
of attention (PoA) scores over time (PoA is the sum
of mean choice reaction time, simple reaction time
and digit vigilance reaction time scores), (iii) to
determine whether the PIGD phenotype is asso-
ciated with changes in PoA and (iv) to establish
whether baseline PoA is associated with subse-
quent cognitive decline. We hypothesised first, that
baseline motor phenotype would predict impair-
ment in attention (with PIGD type associated with
greater deficits) and second, that greater impair-
ment in baseline measures of attention would
independently predict cognitive decline.
 Downloaded from http://jnnp.bmj.com/ on February 8, 2015 - Published by group.bmj.com
METHODS
Recruitment of patients with PD
Consecutive patients were recruited over 24 months from local
dementia and movement disorder clinics in Newcastle upon
Tyne and Sunderland. Eighty-one subjects with a provisional
diagnosis of PD were screened for the study. Thirty-nine of
these patients with PD agreed to participate, met Queen Square
Brain Bank criteria18 and were aged 65 years or over. Subjects
were only included if taking L-dopa monotherapy and no other
medications for treatment of their PD. All subjects were
assessed clinically as not having any form of dementia (on the
basis of DSM-IV criteria) and scored more than 24 on the
MMSE.19 On entry into the study, no subjects were taking
antipsychotic medications or cholinesterase inhibitors. One
subject was receiving an anticholinergic agent at baseline. Of
the 42 subjects with PD who were not recruited, 28 refused
participation, one was too frail, one initially agreed but then did
not complete the baseline assessment and 12 others did not
meet the study criteria.
All subjects were assessed at entry to the study (baseline) and
annually thereafter for 3 years. The 2 year follow-up visit and
3 year visit were performed a mean of 24 (SD 0.4) months and
36 (1.16) months after the baseline visit, respectively.
At baseline, subjects were assessed using the Unified
Parkinsons Disease Rating Scale (UPDRS, activities of daily
living, part II and motor impairments, part III).20 Items derived
from these two subscales permit determination of PD subtype,
according to the method proposed by Jankovic and colleagues.21
In the present report, motor phenotypes were defined at
baseline as either PIGD or non-PIGD (including 18 tremor
dominant and six indeterminate subtypes). At baseline and
follow-up assessments, all subjects underwent cognitive assess-
ments, including the MMSE and CAMCOG.22 Depression was
assessed using the Geriatric Depression Scale 15 (GDS-15).23 At
baseline and follow-up, subjects underwent CDR computer
based attentional tasks that included assessment of simple and
choice reaction times and digit vigilance (all measured in
milliseconds), as described by Wesnes and colleagues.24 PoA
was calculated by summing the choice reaction time, simple
reaction time and digit vigilance reaction time scores; higher
scores represent worse attentional performance and reflect how
well a person can maintain concentration to a particular
task.24 25 Subjects were re-evaluated clinically at each time point
for the development of dementia; diagnosis was made on the
Table 1 Baseline demographic, cognitive and motor characteristics
Non-PIGD group
No of cases at baseline 23
Age (years) 74.5 (5.8)
Sex (men: women)* 17:6
Duration of PD symptoms at baseline (years) 3.8 (4.5)
Baseline UPDRS III 23.9 (8.8)
Baseline MMSE 26.7 (2.3)
Baseline CAMCOG 90.8 (6.5)
Baseline PoA (ms) 1377 (247)
Baseline GDS 4.1 (3.4)
On levodopa at baseline 15
Baseline levodopa dose (mg) 353.3 (235.6)
*Overall preponderance of men over women in the study (p = 0.008).
Means (SD) are displayed where appropriate. MannWhitney U tests were used for comparison unless stated otherwise.
CAMCOG, Cambridge Cognitive Examination; GDS, Geriatric Depression Scale; MMSE, Mini-Mental State Examination; NA, not
applicable; PD, Parkinsons disease; PIGD, postural instability gait difficulty; PoA, power of attention; UPDRS III, Unified Parkinsons
Disease Rating Scale III.
J Neurol Neurosurg Psychiatry 2008;79:13181323. doi:10.1136/jnnp.2008.147629
Research paper
basis of DSM-IV criteria and an MMSE score (24. The study
received approval from the local research ethics committee.
Statistical analysis
SPSS (V.15.0) was used for all statistical analyses. For baseline
variables, comparisons were made between subjects with and
without the PIGD motor phenotype. Baseline variables con-
sidered for inclusion were determined a priori on the basis of
previous evidence of their potential effect on cognitive function
in PD and included: age, sex, duration of parkinsonian
symptoms, UPDRS-III scores, MMSE, CAMCOG and GDS-15.
In addition, baseline PoA score was included. Cognitive decline
was determined by taking an average of the annual change in
MMSE and CAMCOG scores in individual subjects; this process
controlled for intersubject baseline variability in cognitive test
scores that might be accounted for by, for example, poor vision
or low premorbid IQ. In addition, to examine for specific effects
of variables on different cognitive domains, the average annual
change in the different subscores of the CAMCOG (orientation;
attention; language comprehension and expression; memory
remote, recent and learning; praxis; calculation; abstract
thinking; and perception) were included in the analyses.
Similarly, an average of the annual change in PoA score was
also calculated for further analysis.
Outcome measures were examined both graphically and
using the ShapiroWilk test to establish whether they were
approximately normally distributed. For comparison of baseline
characteristics, MannWhitney U tests, Students t tests and x2
tests were used where appropriate.
To examine the association between baseline variables and
rates of decline/change in cognitive and attentional functions, a
two stage statistical analysis was carried out: (i) baseline
variables (age, sex, motor phenotype, UPDRS-III baseline score,
CAMCOG baseline score, GDS baseline score, L-dopa daily dose
and baseline PoA score), if not already categorical, were
dichotomised into two groups around the median, and
comparisons of MMSE, CAMCOG, CAMCOG subscores and
PoA annual rate of change were made using MannWhitney U
tests; (ii) baseline variables associated with significant differ-
ences in outcome variables were then entered into backward
stepwise linear regression analyses. Categorical data were coded
0 and 1 (eg, motor phenotype). The distribution of PoA baseline
scores was positively skewed and thus this variable was log
transformed prior to entry into models. Final models were
PIGD group p Value
14 N/A
76.3 (5.4) 0.81 (t test)
10:4 Fisher exact test p = 1.00
5.3 (4.1) 0.14
27.8 (12.4) 0.30
26.6 (1.8) 0.62
86.9 (6.1) 0.04*
1443 (253) 0.27
3.4 (2.0) 0.89
13 Fisher exact test p = 0.11
526.9 (374.0) 0.21
1319
 Downloaded from http://jnnp.bmj.com/ on February 8, 2015 - Published by group.bmj.com

Research paper

Table 2 Bivariate analysis of the effect of baseline variables on MMSE decline, CAMCOG decline and annual change in PoA
MMSE CAMCOG PoA
Variable (decline/year) p Value (decline/year) p Value (increase/year) p Value

Age (years)
,74 0.60 (1.45) 0.17 1.16 (4.25) 0.92 127.79 (391.34) 0.66
>74 20.06 (1.35) 1.04 (2.92) 79.70 (199.28)
Sex
Men 0.41 (1.57) 0.11 1.76 (3.85) 0.06 140.43 (347.32) 0.06
Women 20.17 (0.82) 20.70 (1.83) 3.26 (82.19)
Motor phenotype
Non-PIGD 20.28 (1.20) 0.002* 20.15 (2.23) 0.02* 11.36 (93.44) 0.01*
PIGD 1.14 (1.34) 3.14 (4.44) 244.03 (443.40)
Duration of PD symptoms at baseline (years)
,3.6 20.39 (1.10) 0.01* 0.17 (2.24) 0.34 67.81 (223.47) 0.35
>3.6 0.81 (1.45) 1.88 (4.31) 136.15 (367.77)
Baseline UPDRS III
,24 0.03 (0.88) 0.29 0.07 (2.34) 0.23 28.37 (81.77) 0.001*
>24 0.43 (1.73) 1.87 (4.17) 185.09 (377.35)
Baseline CAMCOG
,91 0.68 (1.71) 0.01* 2.07 (4.30) 0.19 181.86 (392.23) 0.03*
>91 20.24 (0.76) 20.05 (2.05) 8.52 (80.49)
Baseline GDS-15
,3 0.08 (1.10) 0.19 0.48 (3.13) 0.37 146.07 (430.57) 0.96
>3 0.38 (1.62) 1.52 (3.86) 75.03 (190.18)
Levodopa dose (mg)
,300 20.38 (0.79) 0.03* 20.88 (1.71) 0.03* 6.08 (116.20) 0.04*
>300 0.49 (1.54) 1.83 (3.82) 145.16 (350.02)
Power of attention at baseline (ms)
,1344 20.27 (1.63) 0.008* 0.20 (3.92) 0.04* 25.72 (71.76) 0.28
>1344 0.71 (1.05) 1.86 (3.15) 185.15 (421.09)
Means (SD) are displayed where appropriate.
CAMCOG, Cambridge Cognitive Examination; GDS, Geriatric Depression Scale; MMSE, Mini-Mental State Examination; PD, Parkinsons disease; PIGD, postural instability gait
difficulty; PoA, power of attention; UPDRS III, Unified Parkinsons Disease Rating Scale III.

determined with the stepwise exclusion of variables if p.0.10.
Independent variable interaction was determined by the
inclusion of centred cross product of variables to regression
models and additionally the presence of interactions was
examined using two way ANOVA to look for any significant
interactions between dichotomised independents variables. In
all regression models there was no evidence of multicollinearity
(variance inflation factor ,2) and no cases exerted undue
influence, as evidenced by Cooks distances (D,1) and
Mahalanobis distances (p,0.001). Standardised residuals were
all ,3 and Durbin Watson values ,2.
RESULTS
Thirty-nine subjects with PD were entered into the study. Over
the course of the 3 year period reported here, five patients with
non-demented PD at baseline developed dementia. Four of these
patients had the PIGD phenotype at baseline compared with one
non-PIGD phenotype at baseline (Fishers exact test, p = 0.06).
Two subjects died during follow-up (one after the first year
assessment, the other after the second year assessment). One of
the patients who died had developed dementia; both had the PIGD
phenotype and thus the rate of change in terms of cognitive decline
and attention in these subjects was derived from a reduced number
of data points. Two subjects who were assessed at baseline were
lost to follow-up and were excluded from the analysis. During the
study period, one subject was initiated on donepezil, two subjects
on quetiapine, four subjects on catechol-O-methyltransferase
inhibitors and two subjects on dopamine agonists.
Baseline characteristics of the PIGD and non-PIGD patients
are shown in table 1. There were no differences between the
groups in terms of age, duration of parkinsonian symptoms,
UPDRS-III scores, MMSE, GDS or PoA scores. Similar propor-
tions of patients in both groups were receiving L-dopa
treatment at baseline and dosages were approximately similar.
There was a suggestion that the PIGD group had slightly lower
CAMCOG scores at baseline (p = 0.04). In addition, there was
an overall preponderance of males over females in the study
(p = 0.008) although gender proportions were similar between
the PIGD and non-PIGD groups.
Mean annual decline in MMSE and CAMCOG scores for the
whole group was 0.26 points and 1.09 points, respectively.
PoA scores increased per year by an average of 103 ms.
Bivariate analysis of the effect of baseline variables on MMSE
decline, CAMCOG decline and change in PoA are shown in
table 2. More rapid MMSE decline was associated with PIGD
phenotype, longer duration of parkinsonian symptoms prior to
baseline, lower baseline CAMCOG, higher L-dopa dose and
higher baseline PoA score (p,0.05). More rapid CAMCOG
decline was associated with PIGD phenotype, higher L-dopa
dose and greater baseline PoA score (p,0.05). A higher annual
increase in PoA score was associated with higher baseline
UPDRS III score, lower baseline CAMCOG score and higher
L-dopa dose (p,0.05). These variables were therefore selected
for inclusion into their respective multiple regression models.
Multivariate analysis shown in table 3 demonstrated that in
the final models, PIGD phenotype and higher baseline PoA
scores were significant predictors of subsequent MMSE and
CAMCOG decline. Addition of the centred cross product of
PIGD phenotype and baseline PoA to models did not
demonstrate any significant interaction effect (MMSE decline,

1320 J Neurol Neurosurg Psychiatry 2008;79:13181323. doi:10.1136/jnnp.2008.147629

 Downloaded from http://jnnp.bmj.com/ on February 8, 2015 - Published by group.bmj.com
Figure 1 Influence of motor phenotype on mean (A) MMSE, (B)
CAMCOG and (C) power of attention over time in PIGD and non-PIGD
patients. CAMCOG, Cambridge Cognitive Examination; MMSE, Mini-
Mental State Examination; PIGD, postural instability gait difficulty.
b = 0.41, p = 0.87; CAMCOG decline b = 2.37, p = 0.72). This
observation was confirmed by a lack of interaction in two way
ANOVA of dichotomised baseline PoA score against PIGD
Table 3 Multivariate regression models for MMSE decline, CAMCOG decline and annual change in PoA
MMSE decline/year
R2 = 0.584, F = 8.79, p = 0.001
CAMCOG decline/year
R2 = 0.290, F = 6.93, p = 0.003
Ln (PoA increase/year)
R2 = 0.130, F = 4.72, p = 0.04
CAMCOG, Cambridge Cognitive Examination; MMSE, Mini-Mental State Examination; PIGD, postural instability gait difficulty; PoA,
power of attention.
J Neurol Neurosurg Psychiatry 2008;79:13181323. doi:10.1136/jnnp.2008.147629
Research paper
phenotype for cognitive decline (MMSE decline, F = 1.13,
p = 0.30; CAMCOG decline, F = 0.19, p = 0.67). Of the variables
entered into the analysis of annual change in PoA score, only
PIGD phenotype significantly predicted a higher annual increase
in PoA scores. The influence of motor phenotype on mean
MMSE, CAMCOG and PoA over time is demonstrated
graphically in fig 1. Of the 17 patients with baseline PoA scores
less than 1344 ms (median baseline score), none developed
dementia by the third year compared with five out of 20
patients (25%) who had PoA scores greater than 1344 ms
(Fishers exact test p = 0.05).
Considering CAMCOG subscore changes, PIGD phenotype
was a significant predictor for decline in orientation (b = 0.64,
p = 0.04) whereas baseline PoA score was a significant predictor
for decline in orientation (b = 1.76, p = 0.04), language expres-
sion (b = 2.03, p = 0.01) and remote memory (b = 0.69,
p,0.001).
DISCUSSION
Consistent with previous studies, this prospective study of
patients with non-demented PD demonstrated that PIGD
motor phenotype was associated with a more rapid cognitive
decline.8 9 Additional findings were that: (i) PIGD motor
phenotype, while not associated with obvious baseline atten-
tional impairment, as measured by PoA, led to a more rapid
decline in attentional function over 3 years (ii) poor baseline
attentional function was associated with a more rapid cognitive
decline.
The rate of decline on the MMSE was low at 0.25 points per
year over the 3 year study period in all 39 subjects; this is
consistent with previous findings by our group over a 2 year
longitudinal study and by others.9 26 In contrast, separation of
the cohort into PIGD and non-PIGD motor phenotypes
demonstrated that the PIGD subgroup had a more marked
cognitive decline of 1.15 points annually on the MMSE,
although this was approximately half the rate of cognitive
decline noted in individuals with PD with established demen-
tia.9 26 This higher rate of cognitive decline in the non-tremor
motor phenotype has previously been reported8 and has been
postulated to relate to non-dopaminergic lesions.5 9 26
We hypothesised that given the putative common neural
substrate mediating attention and postural/gait difficulties, we
would find evidence of poor baseline attentional function
associated with PIGD motor phenotype. Interestingly, while we
noted a slight increase in baseline PoA scores in PIGD patients
compared with non-PIGD patients (mean PoA baseline score:
PIGD patients 1443 (SD 253) ms versus non-PIGD patients 1377
(SD 247) ms), this was not significant. However, over the
Variable b coefficient p Value
Constant 219.66 0.03
PIGD motor phenotype 1.30 0.003
Ln (PoA at baseline) 2.69 0.03
Constant 245.52 0.05
PIGD motor phenotype 3.00 0.008
Ln (PoA at baseline) 6.29 0.05
Constant 4.66 ,0.001
PIGD motor phenotype 1.03 0.04
1321
 Downloaded from http://jnnp.bmj.com/ on February 8, 2015 - Published by group.bmj.com
Research paper
course of the study there was a divergence between PIGD and
non-PIGD subjects in terms of PoA, particularly at years 2 and 3
(see fig 1). We observed that PIGD motor phenotype and poor
baseline attention function were associated with an increased
rate of cognitive decline although these predictors appeared to
be independent of each other.
PIGD motor phenotype in patients with non-demented PD is
likely to be mediated, at least in part, by brainstem lesions
affecting non-dopaminergic systems.5 9 In contrast, attentional
function, while dependent on brainstem activation and ascend-
ing cholinergic systems, is likely to be heavily influenced by the
integrity of neocortical systems, including fronto-subcortical
circuits involving the thalamus. Striatal dopaminergic pathways
may also have a role in inhibitory and attentional function in
patients with PD.27 28 We suggest, therefore, that our observa-
tions may indicate that, initially, there are dissociate processes
mediating attentional dysfunction and PIGD motor phenotype,
but given that they arise from common neurodegenerative
processes (eg, progressive loss of cholinergic function) that with
time there is a convergence of symptoms that is inclusive of
attentional, cognitive and motor impairments. Importantly, the
study findings suggest that early changes in PoA, even the
absence of clear cognitive impairment, may predict incipient
cognitive decline and dementia in PD.
The association of low baseline PoA with decline in
CAMCOG subscores of language expression and remote
memory was consistent with known cognitive deficits in PD
in terms of verbal fluency and recollection of remote
events.8 29 30 In contrast, the lack of any obvious relationship
between either PIGD motor phenotype and baseline PoA score
with the CAMCOG attention subscore was unexpected. The
attention subscore in CAMCOG consists of counting from 20 to
1 and serial 7s; thus while this task will require attention
function, the neural substrate is likely to be widely distributed
with a large element dependent on parietal lobe networks which
are less likely to be affected by the PIGD motor phenotype and
attentional function, as measured by PoA. Overall, however, we
would argue that the observations of a relationship between
PIGD status and PoA scores and various CAMCOG subscores
need to be viewed cautiously given the smaller annual rates of
change (and thus greater variability) compared with the global
CAMCOG score.
There did not appear to be any association between GDS-15
scores and cognitive or attentional function; however, subject
scores were low and the effect of any potential depressive
symptoms on cognitive and attentional function is likely to
have been limited in the present study.
Strengths of the study
Strengths of this study include its prospective design, low
dropout rate, inclusion of patients diagnosed according to
formal diagnostic criteria and verification by consensus agree-
ment. In addition, cognitive changes were assessed using both
the MMSE and CAMCOG. Few subjects were taking drugs at
baseline, or were initiated on at follow-up which could have
affected cognitive and attentional scores.
Weaknesses of the study
Potential weaknesses include the relatively small sample size
and potential recruitment bias (via outpatient clinics rather
than a community based sample, with all patients aged over
65 years), which may limit the generalisability of the results.
Indeed, while the male preponderance in the study is in keeping
1322
with the known gender ratio difference in incidence of PD,31 it
could be argued that the current study findings are applicable to
only men. However, gender ratios were similar when patients
were grouped either by motor phenotype (see table 1) or around
median baseline PoA scores (PoA score ,1344 ms: males 12,
females 6; PoA score .1344 ms: males 15, females 4; p = 0.48)
which suggests that the study findings are applicable to both
sexes.
Although not within the dementia range, baseline CAMCOG
scores for the PIGD group were lower than for the non-PIGD
group suggesting that, even at baseline, these subjects were
more likely to have accelerated cognitive decline.
Whether the relative contribution to subsequent cognitive
decline is primarily caused by the PIGD motor phenotype or to
the degree of pre-existing cognitive impairment is not clear from
the present findings. However, Alves and colleagues10 noted in
their longitudinal study that individuals who developed
incident PIGD motor phenotype had a high risk of subsequent
cognitive decline and dementia. This therefore suggests that the
PIGD motor phenotype is likely to be a predominant
contributor to the observed longitudinal cognitive decline in
the present study.
In addition, a follow-up period of 3 years is comparatively
short and while the current study assumes a linear temporal
progression in cognitive decline, it is recognised that this is
unlikely in clinical practice.32 33
In conclusion, our results suggest that the PIGD motor
phenotype and attentional deficits are associated with a faster
rate of cognitive decline in patients with non-demented PD. We
suggest that subtle deficits in attention and PIGD motor
phenotype in patients with non-demented PD are each
predictive of subsequent cognitive impairment and are poten-
tially additive. This conclusion requires validation in a larger
sample but if robust may help to inform prognosis and future
trials of early interventions in preventing or delaying dementia
in PD.
Funding: This work was supported by the Medical Research Council. The Alzheimer
Research Trust supported ENR.
Competing interests: None.
Ethics approval: The study received approval from the local research ethics
committee.
REFERENCES
1. Schrag A, Jahanshahi M, Quinn N. What contributes to quality of life in patients with
Parkinsons disease? J Neurol Neurosurg Psychiatry 2000;69:30812.
2. Levy G, Tang MX, Louis ED, et al. The association of incident dementia with mortality
in PD. Neurology 2002;59:170813.
3. Marras C, Rochon P, Lang AE. Predicting motor decline and disability in Parkinson
disease: a systematic review. Arch Neurol 2002;59:17248.
4. Aarsland D, Andersen K, Larsen JP, et al. Prevalence and characteristics of
dementia in Parkinson disease: an 8-year prospective study. Arch Neurol
2003;60:38792.
5. Emre M. Dementia associated with Parkinsons disease. Lancet Neurol 2003;2:22937.
6. Hughes TA, Ross HF, Musa S, et al. A 10-year study of the incidence of and factors
predicting dementia in Parkinsons disease. Neurology 2000;54:1596603.
7. Levy G, Schupf N, Tang M-X, et al. Combined effect of age and severity on the risk of
dementia in Parkinsons disease. Ann Neurol 2002;51:7229.
8. Williams-Gray CH, Foltynie T, Brayne CEG, et al. Evolution of cognitive dysfunction
in an incident Parkinsons disease cohort. Brain 2007;130:178798.
9. Burn DJ, Rowan EN, Allan LM, et al. Motor subtype and cognitive decline in
Parkinsons disease, Parkinsons disease with dementia, and dementia with Lewy
bodies. J Neurol Neurosurg Psychiatry 2006;77:5859.
10. Alves G, Larsen JP, Emre M, et al. Changes in motor subtype and risk for incident
dementia in Parkinsons disease. Mov Disord 2006;21:112330.
11. Tiraboschi P, Hansen LA, Alford M, et al. Cholinergic dysfunction in diseases with
Lewy bodies. Neurology 2000;54:40711.
12. Ehrt U, Aarsland D. Psychiatric aspects of Parkinsons disease. Curr Opin Psychiatry
2005;18:33541.
J Neurol Neurosurg Psychiatry 2008;79:13181323. doi:10.1136/jnnp.2008.147629
 Downloaded from http://jnnp.bmj.com/ on February 8, 2015 - Published by group.bmj.com

Research paper

13. Litvan I, Mohr E, Williams J, et al. Differential memory and executive functions in
demented patients with Parkinsons and Alzheimers disease. J Neurol Neurosurg
Psychiatry 1991;54:259.
14. Bronnick K, Ehrt U, Emre M, et al. Attentional deficits affect activities of daily living
in dementia-associated with Parkinsons disease. J Neurol Neurosurg Psychiatry
2006;77:113642.
15. Dubois B, Pilon B, Lhermitte F, et al. Cholinergic deficiency and frontal dysfunction in
Parkinsons disease. Ann Neurol 1990;28:11721.
16. Woollacott M, Shumway-Cook A. Attention and the control of posture and gait: a
review of an emerging area of research. Gait Posture 2002;16:114.
17. OShea S, Morris ME, Iansek R. Dual task interference during gait in people with
Parkinson disease: effects of motor versus cognitive secondary tasks. Phys Ther
2002;82:88897.
18. Gibb WR, Lees AJ. The relevance of the Lewy body to the pathogenesis of idiopathic
Parkinsons disease. J Neurol Neurosurg Psychiatry 1988;51:74552.
19. Folstein MF, Folstein SE, McHugh PR. Mini-mental state. A practical method for
grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12:18998.
20. Fahn S, Elton R, and members of the UPDRS development committee. Unified
Parkinsons Disease Rating Scale. In: Fahn S, Mardson C, Goldstein M, et al, eds.
Recent developments in Parkinsons disease. New York: Macmillan, 1987:15363.
21. Jankovic J, McDermott M, Carter J, et al. Variable expression of Parkinsons
disease: a base-line analysis of the DATATOP cohort. The Parkinson Study Group.
Neurology 1990;40:152934.
22. Roth M, Tym E, Mountjoy CQ, et al. CAMDEX. A standardised instrument for the
diagnosis of mental disorder in the elderly with special reference to the early
detection of dementia. Br J Psychiatry 1986;149:698709.
23. Meara J, Mitchelmore E, Hobson P. Use of the GDS-15 geriatric depression scale as
a screening instrument for depressive symptomatology in patients with Parkinsons
disease and their carers in the community. Age Ageing 1999;28:358.
24. Wesnes KA, McKeith IG, Ferrara R, et al. Effects of rivastigmine on cognitive
function in dementia with Lewy bodies: a randomised placebo-controlled international
study using the cognitive drug research computerised assessment system. Dement
Geriatr Cogn Disord 2002;13:18392.
25. Emre M, Aarsland D, Albanese A, et al. Rivastigmine for dementia associated with
Parkinsons disease. N Engl J Med 2004;351:250918.
26. Aarsland D, Andersen K, Larsen JP, et al. The rate of cognitive decline in Parkinson
disease. Arch Neurol 2004;61:190611.
27. Barnes J, Boubert L. Executive functions are impaired in patients with
Parkinsons disease with visual hallucinations. J Neurol Neurosurg Psychiatry
2008;79:1902.
28. Nieoullon A. Dopamine and the regulation of cognition and attention. Prog Neurobiol
2002;67:5383.
29. Leplow B, Dierks C, Herrmann P, et al. Remote memory in Parkinsons disease and
senile dementia. Neuropsychologia 1997;35:54757.
30. Sagar HJ, Cohen NJ, Sullivan EV, et al. Remote memory function in Alzheimers
disease and Parkinsons disease. Brain 1988;111:185206.
31. Taylor KSM, Cook JA, Counsell CE. Heterogeneity in male to female risk for
Parkinsons disease. J Neurol Neurosurg Psychiatry 2007;78:9056.
32. Ballard C, OBrien J, Swann A, et al. One year follow-up of parkinsonism in dementia
with Lewy bodies. Dement Geriatr Cogn Disord 2000;11:21922.
33. Helmes E, Bowler JV, Merskey H, et al. Rates of cognitive decline in Alzheimers
disease and dementia with Lewy bodies. Dement Geriatr Cogn Disord 2003;15:6771.

Quality & Safety in Health Care

Quality & Safety in Health Care is a leading international peer-review journal in the growing area of
quality and safety improvement. It provides essential information for those wanting to reduce harm and
improve patient safety and the quality of care. The journal reports and reflects research, improvement
initiatives and viewpoints and other discursive papers relevant to these crucial aims with contributions
from researchers, clinical professionals and managers and experts in organisational development and
behaviour.

qshc.bmj.com

J Neurol Neurosurg Psychiatry 2008;79:13181323. doi:10.1136/jnnp.2008.147629 1323

 Downloaded from http://jnnp.bmj.com/ on February 8, 2015 - Published by group.bmj.com

Poor attentional function predicts cognitive

decline in patients with non-demented
Parkinson's disease independent of motor
phenotype
J-P Taylor, E N Rowan, D Lett, J T O'Brien, I G McKeith and D J Burn

J Neurol Neurosurg Psychiatry 2008 79: 1318-1323 originally published

online June 27, 2008
doi: 10.1136/jnnp.2008.147629

Updated information and services can be found at:

http://jnnp.bmj.com/content/79/12/1318

These include:

References This article cites 32 articles, 16 of which you can access for free at:
http://jnnp.bmj.com/content/79/12/1318#BIBL

Email alerting Receive free email alerts when new articles cite this article. Sign up in the
service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections

Collections Memory disorders (psychiatry) (1254)
Dementia (911)

Notes

To request permissions go to:

http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to:

http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to:

http://group.bmj.com/subscribe/