Table 2: Factors associated with recurrent falls in Parkinson’s disease.

First author
year
study aim
Participant numbera
Tested ON or OFF
Disease severityReporting periodClassification of participantsNumber per falls classificationVariables examinedAnalysesResults

Allcock et al.
2009 [15]

Determine whether measures of attention were associated with falls

164
(176)

 OFF


Unclear





12 months





0 falls
1 fall
>1 fall



61 (37%)
32 (20%)
71 (43%)



Cognitive impairment
Demographics
Disease severity
Fall history
Nonmotor impairments
PD medications
Other medications
Negative binomial regressionSignificant explanatory variables explaining fall frequency
(i) Disease severity (UPDRS)
(ii) Dopamine agonists
(iii) Cognitive impairment
 (a) Power of attention
 (b) Cognitive reaction time
 (c) Reaction time variability
(iv) Fall history

Bloem et al.
2001 [2]

Identify risk factors associated with falls and prediction of falls, particularly in relation to balance and gait
59
(61)

ON

Mild-moderately severe



6 months



0-1 fall
>1 fall


44 (75%)
15 (25%)


Activities of daily living
Demographics
Disease duration
Disease severity
Fall history
Fear of falling
Medications
Mobility and use of aids
Motor impairments
Multiple task performance
Stepwise forward logistic regressionRecurrent fallers best predicted by the following
(i) Disease severity (H&Y)
(ii) Fall history

Camicioli and Majumdar 2010 [23]

Identify risk factors associated with falls, with a focus on cognitive impairment
52
(52)

ON

Mild-moderate



12 months



≥1 fall
>1 fall


21 (40%)
15 (29%)


Cognitive impairment
Demographics
Disease severity
Fall history
Fear of falling
Gait parameters
Motor impairments
Nonmotor impairments
PD medications
Univariate analysisFactors associated with an increased risk of recurrent falls
(i) Cognitive impairment (CCDRSum)
(ii) Fall history
(iii) Disease severity (H&Y)
(iv) Freezing (UPDRS item)

Foreman et al.
2011 [28]

Examine the Functional
Gait Assessment, the pull test, and the timed up and go and their relation to falls
36
(36)

OFF and ON

Mild-moderately severe



≥6 months



0-1 fall
>1 fall


14 (39%)
22 (61%)


Demographics
Disease duration
Disease severity
Mobility
Motor impairments
Receiver operating characteristic curve Interpretation of performance when OFF provided more accurate prediction of fall status than the ON condition
Between-group comparisons
Compared to single + nonfallers, recurrent fallers had the following
(i) Worse Functional Gait Assessment scores when ON and when OFF
(ii) Slower timed up and go when OFF

Mak and Pang 2009 [33]

Examine whether
fear of falling could independently predict recurrent
falls




70
(72)

ON





Moderate







12 months







0-1 fall
>1 fall






55 (79%)
15 (21%)






Demographics
Disease duration
Disease severity
Fall history
Fear of falling
Medications
Mobility
Nonmotor impairments
Stepwise discriminant analysisFor predicting future recurrent fallers
(i) Fall history strongest predictor
(ii) UPDRS motor score and fear of falling (ABC) remain significant after adjusting for fall history
Receiver operating characteristic curveFor identifying recurrent fallers
(i) ABC cut-off score of 69 (sensitivity 93%, specificity 67%) and UPDRS motor score of 32 (sensitivity 47%, specificity 94%) provide the best combination
Between-group comparisonsCompared to single + nonfallers, recurrent fallers had the following
(i) Increased disease severity (H&Y)
(ii) Higher UPDRS motor scores
(iii) Increased fear of falling

Mak and Pang 2010 [18]

Compare fall characteristics between single and recurrent fallers
72
(74)
  
ON

Mild-moderate




12 months




0 falls
1 fall
>1 fall


47 (65%)
12 (17%)
13 (18%)


21 variables including
Anthropometrics
Demographics
Disease duration
Disease severity
Fall history
Fear of falling
Habitual physical activity
Mobility and use of aids
Motor impairments
Nonmotor impairments
PD medications
Between-group comparisonsCompared to single fallers, recurrent fallers had the following
(i) More previous falls
(ii) Increased PD motor impairments (UPDRS)
(iii) Reduced walking capacity (6 MWD)
(iv) Increased use of walking aids
(v) Reduced speed of sit-to-stand
(vi) Increased fear of falling (ABC)
(vii) A higher proportion of falls occurring indoors at home as opposed to outdoors

Matinolli et al.
2011 [34]




Identify balance and
mobility related risk factors for recurrent falling




125
(125)
  



ON





Mild-moderate




  






24 months




  






0-1 fall
>1 fall




0 falls
1 fall
2–5 falls
6–10 falls
11–100 falls
>208 falls
66 (53%)
59 (47%)




46 (37%)
20 (16%)
22 (17%)
16 (13%)
15 (12%)
6 (5%)
Comorbidities
Cognitive impairment
Demographics
Disease severity
Fall history
Fear of falling
Habitual physical activity
Mobility and use of aids
Motor impairments
Nonmotor impairments
Other medications
PD medications
Forward stepwise regressionSignificant risk factors in the final multivariable model predicting recurrent falls
(i) Fall history
(ii) Disease severity (UPDRS II)
Between-group comparisons









Compared to single + nonfallers, recurrent fallers had the following
(i) Longer disease duration
(ii) Increased disease severity (H&Y and UPDRS ADL score, motor score and total)
(iii) Presence of freezing of gait
(iv) More falls unrelated to freezing of gait (UPDRS item 13)
(v) Experienced recent falls
(vi) Higher levodopa dose
(vii) Decreased physical activity
(viii) Reduced mobility (slowed walking speed and TUG)
(ix) Increased use of walking aids
(x) Increased postural sway

aParticipant number-number reported (number recruited).
NR: not reported; UPDRS: Unified Parkinson’s Disease Rating Scale; H&Y: Hoehn and Yahr stage; CCDRSum: Caregiver-rated Clinical Dementia Rating Scale; ABC: Activities-Specific Balance Confidence Scale; 6MWD: 6-minute walk distance; ADL: activities of daily living; TUG: timed up and go.