Falling in the elderly: Do statistical models matter for performance criteria of fall prediction? Results from two large population-based studies

Postural control among older adults with fear of falling and chronic low back pain

OBJECTIVE

Altered Postural control could increase the risk of falling in older adults. Factors such as low back pain and fear of falling can be contributing factors to postural control instability. This study aimed to investigate the effect of chronic low back pain (CLBP) and fear of falling (FOF) on postural control of older adults.

METHOD

Forty-one older adults were included (27 LBP and 14 control). Among the participants, 22 people had high FOF, and 19 had low FOF based on Falls efficacy scale cut-off of ≥ 26. For postural control evaluation Center of pressure parameters (COP) of Standard deviation (Sd) of velocity, Sd of amplitude, path length and mean velocity in both Medial-lateral (ML) and Anterior-Posterior (AP) directions were measured. Mixed-model anova with two between group factor (Health status; with and without CLBP, and with high and low FES-I groups) and one within factor postural condition (four conditions with and without vision and Achill tendon vibration) was used.

RESULT

No significant interaction between groups (health status and FES-I) and group with condition (health status and condition or FES-I and condition) was observed for all COP parameters in both AP and ML direction. There was main effect of FES-I for all COP parameters in ML direction, with greater Sd of velocity, Sd of amplitude, path length and mean velocity in older adults with high FES-I compared to low FES-I in the ML direction.

CONCLUSION

High levels of FOF influenced static postural control in the ML direction. Therefore, paying attention to the lateral stability of older adults is of great importance.

Association Between Falls and Social Frailty in Community-Dwelling Older Japanese Adults

OBJECTIVES

This study aimed to investigate the association between falls and social frailty and its components among older Japanese adults.

METHODS

This is a cross-sectional study. Participants were categorized into 3 groups based on the number of falls in the past year: no fall (none), a single fall (occasional), and more than one fall (recurrent). The participants who met 2 or more of the following criteria were defined as socially frail: living alone, going out less frequently compared with the previous year, rarely visiting friends, feeling unhelpful to friends or family, and not talking with someone daily.

RESULTS

A total of 4,495 older Japanese adults living in a community analyzed in this study (51.0% women). Of the participants in this study, 3,851 (85.7%) were categorized as none, 443 (9.9%) as occasional, and 201 (4.5%) as recurrent. The proportion of participants considered socially frail was 11.5% in this study. Recurrent falls were associated with social frailty, even after adjusting for covariates (odds ratio [OR]: 1.49; 95% confidence interval [CI]: 1.01-2.19). The experience of recurrent falls was associated with the following components: "feeling unhelpful to friends and family" (OR: 1.62; 95% CI: 1.14-2.31) and "going outside less frequently compared with last year" (OR: 1.57; 95% CI: 1.06-2.31).

DISCUSSION

Among older Japanese adults, recurrent falls were associated with social frailty and with 2 of its components in particular: social roles and social participation. Future longitudinal studies should be conducted to gain insight into any causal relationships between these variables.

Falls prevention in older people and the role of nursing

Falls among older individuals pose a significant public health challenge globally, impacting both individual wellbeing and healthcare systems. This article examines the importance of falls prevention in older people and the pivotal role of nursing in this domain. It presents statistics indicating the high prevalence of falls among older adults, highlighting their substantial impact on morbidity, mortality and healthcare costs. Furthermore, it discusses the multifactorial nature of fall risk factors, including age-related changes, chronic health conditions, medication use, impaired mobility, sensory deficits and environmental hazards. Nursing interventions encompass comprehensive assessments, personalised care plans, patient education and advocacy efforts aimed at reducing fall risks and enhancing safety. By addressing intrinsic and extrinsic factors contributing to falls, nurses contribute significantly to improving the quality of life for older adults and reducing the economic burden associated with fall-related injuries.

ER2 risk levels and their association with incident falls, their recurrence and post-fall fractures in older women: Results of the EPIDOS study

BACKGROUND

"Emergency Room Evaluation and Recommendations" (ER2) is a validated clinical tool which stratifies the risk of the occurrence of adverse outcomes in three levels (i.e., low, moderate and high) in older people attending emergency departments. This study examines the association of ER2 risk levels with incident falls, their recurrence and post-fall fractures in older community women.

METHODS

7147 participants of the EPIDémiologie de l'OStéoporose (EPIDOS) study - an observational population-based cohort study - were selected. ER2 low, moderate and high risk levels were determined at baseline. Incident fall outcomes (i.e., one incident fall without fracture, one incident fall with fracture, ≥2 falls without fracture and ≥ 2 falls with fracture) were collected prospectively every 4 months over a 4-year follow-up period.

RESULTS

The overall incidence of falls was 26.4.%, regardless of their characteristics. ER2 low risk level (hazard ratio (HR) ≤0.80 with P ≤ 0.001) and high risk (HR ≥ 1.26 with P ≤ 0.001) were associated respectively with low and high incident fall outcomes, except for recurrent falls without fracture.

CONCLUSIONS

ER2 low and high risk levels were associated with incident falls outcomes in EPIDOS participants, suggesting that the ER2 tool may be useful for stratifying the risk of falls in the older population.

A comprehensive review of machine learning algorithms and their application in geriatric medicine: present and future

The increasing access to health data worldwide is driving a resurgence in machine learning research, including data-hungry deep learning algorithms. More computationally efficient algorithms now offer unique opportunities to enhance diagnosis, risk stratification, and individualised approaches to patient management. Such opportunities are particularly relevant for the management of older patients, a group that is characterised by complex multimorbidity patterns and significant interindividual variability in homeostatic capacity, organ function, and response to treatment. Clinical tools that utilise machine learning algorithms to determine the optimal choice of treatment are slowly gaining the necessary approval from governing bodies and being implemented into healthcare, with significant implications for virtually all medical disciplines during the next phase of digital medicine. Beyond obtaining regulatory approval, a crucial element in implementing these tools is the trust and support of the people that use them. In this context, an increased understanding by clinicians of artificial intelligence and machine learning algorithms provides an appreciation of the possible benefits, risks, and uncertainties, and improves the chances for successful adoption. This review provides a broad taxonomy of machine learning algorithms, followed by a more detailed description of each algorithm class, their purpose and capabilities, and examples of their applications, particularly in geriatric medicine. Additional focus is given on the clinical implications and challenges involved in relying on devices with reduced interpretability and the progress made in counteracting the latter via the development of explainable machine learning.

An analysis of the associated factors for falls, recurrent falls, and fall-related injuries among the older adults in senior Chinese apartments: A cross-sectional study

Older adults living in care facilities such as senior apartments may experience falls and severe falls (i.e., fall-related injuries or falls ≥2 times), which are associated with multiple risk factors. However, there are few studies on falls among older adults in senior Chinese apartments. The purpose of our study is to investigate the current situation of falls among older adults in senior apartments and analyze the related factors of falls and severe falls, to assist agency workers in identifying older adults who are at high risk of falls and reducing fall occurrence and fall injuries.

Mediation of pain in the association of sleep problems with falls among older adults in India

Body pain, sleep problems and falls are commonly reported among the elderly population. This study aimed to explore the mediating role of pain in the association of sleep problems with fall-outcomes (falls, fall-injury, and multiple falls) among older adults. Cross-sectional data from the baseline survey of Longitudinal Aging Study in India (LASI), 2017-18 were used. The total sample size for the study was 28,285 older adults aged 60 years and above. Falls and fall-related injuries among older adults in the last two years were self-reported. The Jenkins Sleep Scale (JSS-4) was used to assess sleep problems while pain was assessed using questions on whether respondents reported that they were troubled by pain and they required some form of medication or treatment for the relief of pain. Multivariable logistic regression and mediation analyses were conducted to fulfill the study objectives. While 13% older adults suffered from sleep problems, 38.83% were troubled with pain. Additionally, 12.63%, 5.64% and 5.76% older adults reported falls, fall-injury and multiple falls respectively. Older adults who suffered from sleep problems had higher odds of falls [adjusted odds ratio (aOR): 1.43, confidence interval (CI): 1.30-1.58], fall-injuries, [aOR:1.50,CI:1.30-1.73] and multiple falls [aOR:1.41,CI:1.24-1.62]. Similarly, older adults who were troubled with pain were more likely to report falls [aOR:1.80, CI:1.67-1.95], fall-injuries [aOR:1.66, CI:1.48-1.87] and multiple falls [aOR:1.90,CI:1.69-2.12]. The percent of the mediated effect of pain when examining the association between sleep problems and fall outcomes were reported to be 17.10%, 13.56% and 18.78% in case of falls, fall-injuries and multiple falls respectively. The current study finds evidence that pain mediates the association of sleep problems and falls, fall-injuries, and multiple falls among older Indian adults. Both sleep problems and pain are modifiable risk factors that need attention for fall prevention strategies.

Establishing the minimal clinically important difference of the EQ-5D-3L in older adults with a history of falls

PURPOSE

Establish the minimal clinically important difference (MCID) of a health-related quality of life (HRQoL) measure-the EuroQol EQ-5 Dimensions-3 Level (EQ-5D-3L)-in older adults with a history of falls.

METHODS

This study is a secondary analysis of 255 complete cases who were enrolled in a 12-month randomized controlled trial (NCT01029171; NCT00323596); participants were randomized to the Otago Exercise Program (OEP; n = 126/172; Age:81.2 ± 6.2 years; 60.3% Female) or control (CON; n = 129/172; Age:81.7 ± 5.7 years; 70.5% Female). Participants completed the EQ-5D-3L and Visual Analogue Scale (VAS) at baseline and 1-year. The VAS was associated with HRQoL and was the health status anchor (VAS minimal improvement = 7 to 17, maximal improvement ≥ 18, minimal decline =  - 7 to - 17, maximal decline ≤  - 18 points). We used four distinct approaches to estimate MCID ranges: (1) anchor-based change differences of the EQ-5D-3L (1-year minus baseline); (2) anchor-based beta coefficients from ordinary least squares regressions (OLS); (3) anchor-based receiver operating characteristic (ROC), and 4) distribution-based standard deviation and standardized effect size of 0.5.

RESULTS

EQ-5D-3L MCID ranges for minimal improvements (OEP = 0.028 to 0.059; CON = 0.007 to 0.051), maximal improvements (OEP = 0.059 to 0.090; CON = 0.051 to 0.090), minimal declines (OEP =  - 0.029 to - 0.105; CON =  - 0.015 to - 0.051), and maximal declines (OEP =  - 0.018 to - 0.072; CON =  - 0.018 to - 0.082) were established using change difference, OLS, and distribution-based methods. The ROC area under the curve was poor, thus, it was not used to estimate the MCID.

CONCLUSIONS

Our results will assist in the interpretation of changes in HRQoL, as measured by the EQ-5D-3L, in older adults with a history of falls.

Using conditional inference forests to examine predictive ability for future falls and syncope in older adults: Results from The Irish Longitudinal Study on Ageing

BACKGROUND

The extent to which gait and mobility measures predict falls relative to other risk factors is unclear. This study examined predictive accuracy of over 70 baseline risk factors, including gait and mobility, for future falls and syncope using conditional inference forest models.

METHODS

Data from three waves of The Irish Longitudinal Study on Ageing (TILDA), a population-based study of community-dwelling adults aged ≥50 years were used (n=4,706). Outcome variables were recurrent falls, injurious falls, unexplained falls and syncope occurring over four year follow-up. Predictive accuracy was calculated using 5 fold cross-validation; as there was class imbalance, the algorithm was trained using undersampling of the larger class. Classification rate, area under the receiver operating characteristic curve (AUROC) and area under the precision recall curve (PRAUC) assessed predictive accuracy.

RESULTS

Highest overall accuracy was 69.7% for recurrent falls in 50-64 year olds. AUROC and PRAUC were ≤0.69 and ≤0.39 respectively for all outcomes indicating low predictive accuracy. History of falls, unsteadiness while walking, fear of falling, mobility, medications , mental health and cardiovascular health and function were the most important predictors for most outcomes.

CONCLUSIONS

Conditional inference forest models using over 70 risk factors resulted in low predictive accuracy for future recurrent, injurious and unexplained falls and syncope in community-dwelling adults. Gait and mobility impairments were important predictors of most outcomes but did not discriminate well between fallers and non-fallers. Results highlight the importance of multifactorial risk assessment and intervention and validate key modifiable risk factors for future falls and syncope.

Comparison of factors influencing fall recurrence in the young-old and old-old: a cross-sectional nationwide study in South Korea

Background

Recurrent falls are a concerning problem in the elderly. Elderly people aged > 65 years who are prone to fall often require medical treatment for severe fall-related injuries, which is associated with a substantial financial burden. Therefore, this study aimed to identify factors related to recurrent falls in the community-dwelling young-old (65–74 years old) and old-old (≥ 75 years) in South Korea.

Methods

This study used a cross-sectional, correlation design. Data from the 2017 National Survey of Older Koreans were used, and 5,838 young-old and 4,205 old-old elderly people were included in the analysis. The questionnaire included general characteristics, fall experience, physical status, mental status, and presence of chronic diseases. The data were analyzed using the chi-square test, one-way analysis of variance, and logistic regression analysis.

Results

In the young-old elderly people, limitations in activities of daily living (p < .001), use of visual aids (p = .002), cognitive function (p < .001), presence of suicidal ideations (p = .005), number of chronic diseases (p < .001), and number of prescribed medications used (p = .006) associated with fall recurrence. In the old-old elderly people, having a spouse (p = .034), being a beneficiary of the National Basic Livelihood Security System (p = .025), less exercise (p = .003), limitations in activities of daily living (p < .001), visual aid use (p = .002), presence of suicidal ideations (p = .015), number of chronic diseases (p < .001), and presence of Parkinson's disease (p < .001) associated with fall recurrence.

Conclusions

This study identified differences in factors related to fall recurrence between the young-old and old-old elderly. The results of this study indicate that it is necessary to implement an intervention program to prevent fall recurrence by age group in consideration of the risk factors for fall recurrence in each elderly people group.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12877-022-03172-7.

Oral Health and Related Quality of Life in Older People: A Systematic Review and Meta-Analysis

Background

Of the most important implications and complaints in the elderly group of the population, is oral and dental health problems. This study aimed to assess oral health- related quality of life in older people.

Methods

To data collection, databases were searched including PubMed, EMBASE, Scopus, SID, MagIran, Cochrane Central Register of Controlled Trials and scholar google The keywords were "older adults", "Geriatric" Elderly", "Older", "Aged", "Ageing", "Oral health", "Oral hygiene" and "Quality of life", "QOL. For manual searching, several specialized journals of related scope as well as the finalized articles' reference list were searched. Studies from 1st Jan 2000 to 30^th Jan 2017 were included. Studies were subjected to meta-analysis to calculate indexes, using CMA:2 (Comprehensive Meta-Analysis) software.

Results

Totally, 3707 articles were searched that 48 of them were subjected to the oral and dental health-related quality of life in 59 groups of the elderly population with the mean age of 73.57+6.62 in the 26 countries. The obtained percentage values of dental and oral health were 80.2% (0-60), 14.8% (0-12), 16.4% (0-70), 22% (0-14 or 0-59) and 19.2% (0-196) for GOHAI with the additive method, GOHAI with Simple Count Method, OHIP-14 with the additive method, OHIP-14 with Simple Count method and OHIP-49 with additive method indexes, respectively.

Conclusion

The elderly group of the population had no proper oral health-related quality of life. Regarding the importance and necessity of oral and dental health and its effect on general health care in the target group, it is recommended to improve dental hygiene in the mentioned group of population.

Machine learning for predictive data analytics in medicine: A review illustrated by cardiovascular and nuclear medicine examples

The evidence-based medicine allows the physician to evaluate the risk-benefit ratio of a treatment through setting and data. Risk-based choices can be done by the doctor using different information. With the emergence of new technologies, a large amount of data is recorded offering interesting perspectives with machine learning for predictive data analytics. Machine learning is an ensemble of methods that process data to model a learning problem. Supervised machine learning algorithms consist in using annotated data to construct the model. This category allows to solve prediction data analytics problems. In this paper, we detail the use of supervised machine learning algorithms for predictive data analytics problems in medicine. In the medical field, data can be split into two categories: medical images and other data. For brevity, our review deals with any kind of medical data excluding images. In this article, we offer a discussion around four supervised machine learning approaches: information-based, similarity-based, probability-based and error-based approaches. Each method is illustrated with detailed cardiovascular and nuclear medicine examples. Our review shows that model ensemble (ME) and support vector machine (SVM) methods are the most popular. SVM, ME and artificial neural networks often lead to better results than those given by other algorithms. In the coming years, more studies, more data, more tools and more methods will, for sure, be proposed.

The role of impulsivity in neurorehabilitation: A prospective cohort study of a potential cognitive biomarker for fall risk?

BACKGROUND

Executive dysregulation and impulsivity can both predispose individuals to risk-prone actions. Although the risk of falls is well established in people with poor executive function, its association to impulsivity is less clear.

PURPOSE

To describe and assess the prognostic capabilities of the relationship between impulsivity, executive function, functional capability, and falls in the in-patient neurorehabilitation population.

MATERIALS AND METHODS

A prospective cohort study in a 26-bed neurorehabilitation unit in London, recruiting 121 patients, of whom 94 were deemed eligible for inclusion. Cognitive-behavioural assessment was undertaken using the short (16-item) version of the Urgency-Premeditation-Perseverance-Sensation Seeking-Positive Urgency (UPPS) impulsive behaviour scale, and the Trail Making Test (TMT). Patients also underwent a functional assessment at admission and discharge using the UK Functional Independence and Assessment Measure tool (FIM + FAM). The main outcome of interest was falling during an in-patient episode, which are routinely recorded in a computerized registry of adverse incidents.

RESULTS

Measurements of impulsivity (based on the UPPS-Short form) and executive function (based on the Trail Making Test) were not found to be significantly associated with functional improvement, or risk of falling. Predictive modelling experiments demonstrated that neither of the aforementioned results were capable of identifying individuals at risk of falling more accurately than an informed guess.

CONCLUSION

Where impulsivity is present, measurement using structured tools such as the UPPS may be informative to guide individualized rehabilitation programmes; however, its usefulness as the basis of risk prediction models for falls is less likely given the results of this study.

Risk factors for recurrent falls in older adults: A systematic review with meta-analysis

Older adults who fall recurrently (i.e., 2 or more falls/year) are at risk of functional decline and mortality. Understanding which risk factors for recurrent falls are most important will inform secondary fall prevention strategies that can reduce recurrent falls risk. Thus, we conducted a systematic review with meta-analysis to determine the relative risk of recurrent falls for different types of falls risk factors. MEDLINE, EMBASE, PsycINFO, and CINAHL databases were searched on April 25, 2019 (Prospero Registration: CRD42019118888). We included peer-reviewed prospective studies which examined risk factors that contributed to recurrent falls in adults aged ≥ 60 years. Using the falls risk classification system of Lord and colleagues, we classified each risk factor into one of the following domains: 1) balance and mobility; 2) environmental; 3) psychological; 4) medical; 5) medication; 6) sensory and neuromuscular; or 7) sociodemographic. We calculated the summary relative risk (RR) for each domain and evaluated the risk of bias and quality of reporting. Twenty-two studies were included in this systematic review and meta-analysis. Four domains predicted recurrent falls: balance and mobility (RR:1.32;95 % CI:[1.10, 1.59]), medication (RR:1.53;95 % CI:[1.11, 2.10]), psychological (RR:1.35;95 % CI:[1.03, 1.78]), and sensory and neuromuscular (RR:1.51;95 % CI:[1.18, 1.92]). Each of these four domains can be viewed as a marker of frailty. The risk of bias was low, and the study quality was high (minimum:19/22). Older adults with markers of frailty are up to 53 % more likely to experience recurrent falls. Strategies that identify and resolve frailty markers should be a frontline approach to preventing recurrent falls.

A systematic review shows no performance benefit of machine learning over logistic regression for clinical prediction models

OBJECTIVES

The objective of this study was to compare performance of logistic regression (LR) with machine learning (ML) for clinical prediction modeling in the literature.

STUDY DESIGN AND SETTING

We conducted a Medline literature search (1/2016 to 8/2017) and extracted comparisons between LR and ML models for binary outcomes.

RESULTS

We included 71 of 927 studies. The median sample size was 1,250 (range 72-3,994,872), with 19 predictors considered (range 5-563) and eight events per predictor (range 0.3-6,697). The most common ML methods were classification trees, random forests, artificial neural networks, and support vector machines. In 48 (68%) studies, we observed potential bias in the validation procedures. Sixty-four (90%) studies used the area under the receiver operating characteristic curve (AUC) to assess discrimination. Calibration was not addressed in 56 (79%) studies. We identified 282 comparisons between an LR and ML model (AUC range, 0.52-0.99). For 145 comparisons at low risk of bias, the difference in logit(AUC) between LR and ML was 0.00 (95% confidence interval, -0.18 to 0.18). For 137 comparisons at high risk of bias, logit(AUC) was 0.34 (0.20-0.47) higher for ML.

CONCLUSION

We found no evidence of superior performance of ML over LR. Improvements in methodology and reporting are needed for studies that compare modeling algorithms.

A new tool for the evaluation of the rehabilitation outcomes in older persons: a machine learning model to predict functional status 1 year ahead

PURPOSE

To date, the assessment of disability in older people is obtained utilizing a Comprehensive Geriatric Assessment (CGA). However, it is often difficult to understand which areas of CGA are most predictive of the disability. The aim of this study is to evaluate the possibility to early predict-1 year ahead-the disability level of a patient using machine leaning models.

METHODS

Community-dwelling older people were enrolled in this study. CGA was made at baseline and at 1 year follow-up. After collecting input/independent variables (i.e., age, gender, schooling followed, body mass index, information on smoking, polypharmacy, functional status, cognitive performance, depression, nutritional status), we performed two distinct Support Vector Machine models (SVMs) able to predict functional status 1 year ahead. To validate the choice of the model, the results achieved with the SVMs were compared with the output produced by simple linear regression models.

RESULTS

218 patients (mean age = 78.01; SD = 7.85; male = 39%) were recruited. The combination of the two SVMs is able to achieve a higher prediction accuracy (exceeding 80% instances correctly classified vs 67% instances correctly classified by the combination of the two linear regression models). Furthermore, SVMs are able to classify both the three categories, self sufficiently, disability risk and disability, while linear regression model separates the population only in two groups (self-sufficiency and disability) without identifying the intermediate category (disability risk) which turns out to be the most critical one.

CONCLUSIONS

The development of such a model can contribute to the early detection of patients at risk of self-sufficiency loss.

The Trail Making test: a study of its ability to predict falls in the acute neurological in-patient population

Objective:

To determine whether tests of cognitive function and patient-reported outcome measures of motor function can be used to create a machine learning-based predictive tool for falls.

Design:

Prospective cohort study.

Setting:

Tertiary neurological and neurosurgical center.

Subjects:

In all, 337 in-patients receiving neurosurgical, neurological, or neurorehabilitation-based care.

Main Measures:

Binary (Y/N) for falling during the in-patient episode, the Trail Making Test (a measure of attention and executive function) and the Walk-12 (a patient-reported measure of physical function).

Results:

The principal outcome was a fall during the in-patient stay ( n = 54). The Trail test was identified as the best predictor of falls. Moreover, addition of other variables, did not improve the prediction (Wilcoxon signed-rank P < 0.001). Classical linear statistical modeling methods were then compared with more recent machine learning based strategies, for example, random forests, neural networks, support vector machines. The random forest was the best modeling strategy when utilizing just the Trail Making Test data (Wilcoxon signed-rank P < 0.001) with 68% (± 7.7) sensitivity, and 90% (± 2.3) specificity.

Conclusion:

This study identifies a simple yet powerful machine learning (Random Forest) based predictive model for an in-patient neurological population, utilizing a single neuropsychological test of cognitive function, the Trail Making test.

Predicting first fall in newly diagnosed Parkinson's disease: Insights from a fall-naïve cohort

BACKGROUND

Falls are common and associated with reduced independence and mortality in Parkinson's disease. Previous research has been conducted on falls-prevalent or advanced disease cohorts.

OBJECTIVE

This study identifies risk factors for first fall for 36 months in a newly diagnosed, falls-naïve cohort.

METHODS

A total of 121 consecutive Parkinson's disease patients were recruited. Falls data were collected prospectively during 36 months from diagnosis via monthly falls diaries and telephone follow-up for 117 participants. Assessment comprised a comprehensive battery of clinical, gait, and cognitive measures. Significant predictors were identified from decision-tree analysis and survival analysis with time to first fall during 36 months as the dependent variable.

FINDINGS

At baseline, 26 (22%) participants reported retrospective falls. At 36 months, the remaining cohort (n = 91) comprised 47 fallers (52%) and 30 (33%) nonfallers and 14 (15%) participants with incomplete diaries. Fallers presented with a significantly higher disease severity, poorer ability to stand on one leg, slower gait speed, increased stance time variability, and higher swing time asymmetry. Median time to first fall was 847 days. Gait speed, stance time, and Hoehn & Yahr III stage emerged as significant predictors of first fall, hazard ratio 3.44 (95% confidence interval [CI] 1.58 to 7.48), 3.31(95% CI 1.40 to 7.80), and 2.80 (95% CI 1.38 to 5.65), respectively. The hazard ratio for risk factors combined was 7.82 (CI 2.80 to 21.84).

CONCLUSIONS

Interventions that target gait deficit and postural control in early Parkinson's disease may limit the potential for first fall. © 2016 International Parkinson and Movement Disorder Society.