Predicting falls in people with Parkinson’s disease: impact of methodological approaches on predictors identified

Lorena R.S. Almeida, Gillian Heller, Natalie E. Allen, Colleen G. Canning, Catherine Sherrington, Guilherme T. Valenca, Jamary Oliveira-Filho, Serene S. Paul

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Background: Previous investigations of falls predictors in people with Parkinson’s disease (PD) have used various statistical methods and categorization of falls outcomes. The impact of methodological differences on falls predictors has not been investigated. Objectives: To describe similarities and differences in predictors modelled for fall rates [negative binomial (NB), Poisson Inverse Gaussian (PIG) and quantile regression] and previously-reported predictors of time to second fall (Cox regression), i.e. past falls, motor fluctuations, disability, levodopa dose and balance impairment. To investigate whether predictors from quantile regression vary across subsets of fallers based on fall frequency. Methods: Participants with PD (n = 229) were followed-up for 12 months. NB and PIG regression were used to determine predictors of fall rates, with the best fitting model reported. Quantile regression was used to determine predictors at higher (62nd, 70th, 80th) percentiles of the falls distribution. Univariate and multivariate analyses were performed. Results: Predictors of fall rates were the same in NB and PIG multivariate models, with the PIG model fitting our data better. Past falls, disability and levodopa dose were associated with fall rates from PIG and quantile regression. Freezing of gait was associated with fall rates from PIG regression. Disease severity predicted less (70th percentile, approximately 2–4) and more (80th percentile, approximately ≥ 5) frequent falls, and anteroposterior stability also predicted less frequent falls (p < 0.05), from quantile regression. Conclusions: Not all predictors of time to second fall were predictors of fall rates. Quantile regression revealed some divergent predictors depending on the percentile of fall frequency examined.

LanguageEnglish
JournalAging clinical and experimental research
DOIs
Publication statusE-pub ahead of print - 25 Jul 2019

Fingerprint

Levodopa
Parkinson Disease
Gait
Freezing
Multivariate Analysis

Keywords

  • Falls
  • Parkinson’s disease
  • Prediction
  • Risk factors
  • Statistical models

Cite this

Almeida, Lorena R.S. ; Heller, Gillian ; Allen, Natalie E. ; Canning, Colleen G. ; Sherrington, Catherine ; Valenca, Guilherme T. ; Oliveira-Filho, Jamary ; Paul, Serene S. / Predicting falls in people with Parkinson’s disease : impact of methodological approaches on predictors identified. In: Aging clinical and experimental research. 2019.
@article{3568f6bd63d3415b8d4d94b5caf94cac,
title = "Predicting falls in people with Parkinson’s disease: impact of methodological approaches on predictors identified",
abstract = "Background: Previous investigations of falls predictors in people with Parkinson’s disease (PD) have used various statistical methods and categorization of falls outcomes. The impact of methodological differences on falls predictors has not been investigated. Objectives: To describe similarities and differences in predictors modelled for fall rates [negative binomial (NB), Poisson Inverse Gaussian (PIG) and quantile regression] and previously-reported predictors of time to second fall (Cox regression), i.e. past falls, motor fluctuations, disability, levodopa dose and balance impairment. To investigate whether predictors from quantile regression vary across subsets of fallers based on fall frequency. Methods: Participants with PD (n = 229) were followed-up for 12 months. NB and PIG regression were used to determine predictors of fall rates, with the best fitting model reported. Quantile regression was used to determine predictors at higher (62nd, 70th, 80th) percentiles of the falls distribution. Univariate and multivariate analyses were performed. Results: Predictors of fall rates were the same in NB and PIG multivariate models, with the PIG model fitting our data better. Past falls, disability and levodopa dose were associated with fall rates from PIG and quantile regression. Freezing of gait was associated with fall rates from PIG regression. Disease severity predicted less (70th percentile, approximately 2–4) and more (80th percentile, approximately ≥ 5) frequent falls, and anteroposterior stability also predicted less frequent falls (p < 0.05), from quantile regression. Conclusions: Not all predictors of time to second fall were predictors of fall rates. Quantile regression revealed some divergent predictors depending on the percentile of fall frequency examined.",
keywords = "Falls, Parkinson’s disease, Prediction, Risk factors, Statistical models",
author = "Almeida, {Lorena R.S.} and Gillian Heller and Allen, {Natalie E.} and Canning, {Colleen G.} and Catherine Sherrington and Valenca, {Guilherme T.} and Jamary Oliveira-Filho and Paul, {Serene S.}",
year = "2019",
month = "7",
day = "25",
doi = "10.1007/s40520-019-01281-9",
language = "English",
journal = "Aging clinical and experimental research",
issn = "1594-0667",
publisher = "Springer, Springer Nature",

}

Predicting falls in people with Parkinson’s disease : impact of methodological approaches on predictors identified. / Almeida, Lorena R.S.; Heller, Gillian; Allen, Natalie E.; Canning, Colleen G.; Sherrington, Catherine; Valenca, Guilherme T.; Oliveira-Filho, Jamary; Paul, Serene S.

In: Aging clinical and experimental research, 25.07.2019.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Predicting falls in people with Parkinson’s disease

T2 - Aging clinical and experimental research

AU - Almeida, Lorena R.S.

AU - Heller, Gillian

AU - Allen, Natalie E.

AU - Canning, Colleen G.

AU - Sherrington, Catherine

AU - Valenca, Guilherme T.

AU - Oliveira-Filho, Jamary

AU - Paul, Serene S.

PY - 2019/7/25

Y1 - 2019/7/25

N2 - Background: Previous investigations of falls predictors in people with Parkinson’s disease (PD) have used various statistical methods and categorization of falls outcomes. The impact of methodological differences on falls predictors has not been investigated. Objectives: To describe similarities and differences in predictors modelled for fall rates [negative binomial (NB), Poisson Inverse Gaussian (PIG) and quantile regression] and previously-reported predictors of time to second fall (Cox regression), i.e. past falls, motor fluctuations, disability, levodopa dose and balance impairment. To investigate whether predictors from quantile regression vary across subsets of fallers based on fall frequency. Methods: Participants with PD (n = 229) were followed-up for 12 months. NB and PIG regression were used to determine predictors of fall rates, with the best fitting model reported. Quantile regression was used to determine predictors at higher (62nd, 70th, 80th) percentiles of the falls distribution. Univariate and multivariate analyses were performed. Results: Predictors of fall rates were the same in NB and PIG multivariate models, with the PIG model fitting our data better. Past falls, disability and levodopa dose were associated with fall rates from PIG and quantile regression. Freezing of gait was associated with fall rates from PIG regression. Disease severity predicted less (70th percentile, approximately 2–4) and more (80th percentile, approximately ≥ 5) frequent falls, and anteroposterior stability also predicted less frequent falls (p < 0.05), from quantile regression. Conclusions: Not all predictors of time to second fall were predictors of fall rates. Quantile regression revealed some divergent predictors depending on the percentile of fall frequency examined.

AB - Background: Previous investigations of falls predictors in people with Parkinson’s disease (PD) have used various statistical methods and categorization of falls outcomes. The impact of methodological differences on falls predictors has not been investigated. Objectives: To describe similarities and differences in predictors modelled for fall rates [negative binomial (NB), Poisson Inverse Gaussian (PIG) and quantile regression] and previously-reported predictors of time to second fall (Cox regression), i.e. past falls, motor fluctuations, disability, levodopa dose and balance impairment. To investigate whether predictors from quantile regression vary across subsets of fallers based on fall frequency. Methods: Participants with PD (n = 229) were followed-up for 12 months. NB and PIG regression were used to determine predictors of fall rates, with the best fitting model reported. Quantile regression was used to determine predictors at higher (62nd, 70th, 80th) percentiles of the falls distribution. Univariate and multivariate analyses were performed. Results: Predictors of fall rates were the same in NB and PIG multivariate models, with the PIG model fitting our data better. Past falls, disability and levodopa dose were associated with fall rates from PIG and quantile regression. Freezing of gait was associated with fall rates from PIG regression. Disease severity predicted less (70th percentile, approximately 2–4) and more (80th percentile, approximately ≥ 5) frequent falls, and anteroposterior stability also predicted less frequent falls (p < 0.05), from quantile regression. Conclusions: Not all predictors of time to second fall were predictors of fall rates. Quantile regression revealed some divergent predictors depending on the percentile of fall frequency examined.

KW - Falls

KW - Parkinson’s disease

KW - Prediction

KW - Risk factors

KW - Statistical models

UR - http://www.scopus.com/inward/record.url?scp=85069672298&partnerID=8YFLogxK

U2 - 10.1007/s40520-019-01281-9

DO - 10.1007/s40520-019-01281-9

M3 - Article

JO - Aging clinical and experimental research

JF - Aging clinical and experimental research

SN - 1594-0667

ER -