Dealing with the heterogeneous presentations of freezing of gait: how reliable are the freezing index and heart rate for freezing detection?

Abstract

Background: Freezing of gait (FOG) is an unpredictable gait arrest that hampers the lives of 40% of people with Parkinson's disease. Because the symptom is heterogeneous in phenotypical presentation (it can present as trembling/shuffling, or akinesia) and manifests during various circumstances (it can be triggered by e.g. turning, passing doors, and dual-tasking), it is particularly difficult to detect with motion sensors. The freezing index (FI) is one of the most frequently used accelerometer-based methods for FOG detection. However, it might not adequately distinguish FOG from voluntary stops, certainly for the akinetic type of FOG. Interestingly, a previous study showed that heart rate signals could distinguish FOG from stopping and turning movements. This study aimed to investigate for which phenotypes and evoking circumstances the FI and heart rate might provide reliable signals for FOG detection.

Methods: Sixteen people with Parkinson's disease and daily freezing completed a gait trajectory designed to provoke FOG including turns, narrow passages, starting, and stopping, with and without a cognitive or motor dual-task. We compared the FI and heart rate of 378 FOG events to baseline levels, and to stopping and normal gait events (i.e. turns and narrow passages without FOG) using mixed-effects models. We specifically evaluated the influence of different types of FOG (trembling vs akinesia) and triggering situations (turning vs narrow passages; no dual-task vs cognitive dual-task vs motor dual-task) on both outcome measures.

Results: The FI increased significantly during trembling and akinetic FOG, but increased similarly during stopping and was therefore not significantly different from FOG. In contrast, heart rate change during FOG was for all types and during all triggering situations statistically different from stopping, but not from normal gait events.

Conclusion: When the power in the locomotion band (0.5-3 Hz) decreases, the FI increases and is unable to specify whether a stop is voluntary or involuntary (i.e. trembling or akinetic FOG). In contrast, the heart rate can reveal whether there is the intention to move, thus distinguishing FOG from stopping. We suggest that the combination of a motion sensor and a heart rate monitor may be promising for future FOG detection.

Keywords: Accelerometer; Freezing index; Freezing of gait; Heart rate; Movement disorders; Wearable sensors.

Fig. 1

Overview of the performed gait tasks. Each round started with 10 s of standing at the starting point followed by 360 degree turns in alternating directions with and without a cDT (30 s each, in pseudorandomized order). Next, the participants preceded with the gait trajectory as indicated by the orange arrows, passing between the chair and wall (47 cm wide); passing through the doorway (89 cm wide) into the narrow quarter where they completed a 180 and 360 degree turn in alternating directions; passing between the chairs (54 cm wide); walking straight, making a 180 degree turn and walking back between the chairs and the chair and the wall to the starting point where the gait trajectory was started over. The gait trajectory was performed with a cDT, a mDT or noDT for 90 s each, in a pseudorandomized order. In total, at least four of such rounds of approximately 6 min each were completed and in between each round participants could rest as long as needed. All gait tasks were recorded by three video cameras positioned at strategic locations of the lab. (cDT cognitive dual-task, mDT motor dual-task, noDT no dual-task)

Fig. 2

Overview of the annotated FOG events for each participant, including the participant with Multiple System Atrophy (participant 8). A boxplots of the FOG durations (s) for each participant. The data is clipped at 75 s (3 events were longer than this upper limit). B–D Subdivision of the number of annotated FOG events per type (B), trigger (C), and DT condition (D) for each participant. (FOG Freezing of Gait; DT dual-task)

Fig. 3

Overall time course of the FI (A), and heart rate (B) for FOG (orange), normal gait events (green), and, stopping (blue). The lines with the shaded areas represent the mean values with 95% confidence intervals over the 14 included participants (n = 14). The heart rate was z-transformed and baseline corrected (− 6 to − 3 s.) to account for individual variances in baseline heart rate and heart rate variability. No z-transformation was applied to the FI because these values are usually evaluated based on absolute values rather than relative values. The boxes indicate the time intervals over which the variables are averaged when exporting to Rstudio: baseline (− 6 to − 3 s), preFOG (− 3 to 0 s), and FOG (0 to 3 s). The figures were created with MATLAB (R2019a) and the Fieldtrip toolbox. (FI Freezing Index, FOG Freezing of Gait)

Fig. 4

Results of the post-hoc analyses of the linear mixed-model analysis of the first model comparing the FI (A–C) (model 1.a), and heart rate (D–F) (model 1.b) for the differences between preFOG and baseline (light orange), and FOG and baseline (dark orange). The point ranges indicate the estimated differences with standard errors of the post-hoc analyses for each FOG type (first column), FOG trigger (second column), and DT condition (third column), but the symbols are only filled when significant interaction effects were found for this factor by time. This means that the panels with the hollow symbols followed the main effects of Table 2. Results of the post-hoc analysis that were significant after p-value correction are indicated with an asterisk (*, < 0.05; **, < 0.005). Figures were created with the ggplot2 package in Rstudio. (FI Freezing Index; FOG Freezing of Gait; noDT no dual-task; cDT cognitive dual-task; mDT motor dual-task)

Fig. 5

Results of the post-hoc analyses of the linear mixed-model analysis of the second model comparing the FI (A–C) (model 2.a), and heart rate change (D–F) (model 2.b) for the differences between FOG and a normal gait event (green), and FOG and stopping (blue). The point ranges indicate the estimated differences with standard errors of the post-hoc analyses for each FOG type (first column), FOG trigger (second column), and DT condition (third column), but the symbols are only filled when a significant interaction effect of this factor was found with condition (i.e. FOG—normal gait event or FOG—stop). This means that the panel with the hollow symbols followed the main effects of Table 2. Results of the post-hoc analysis that were significant after p-value correction are indicated with an asterisk (*, < 0.05; **, < 0.005). Figures were created with the ggplot2 package in Rstudio. (FI Freezing Index, FOG Freezing of Gait, noDT no dual-task, cDT cognitive dual-task, mDT motor dual-task)

Fig. 6

Graphical representation of how a hypothetical new FOG detection method would work by combining a motion sensor (IMU) and a heart rate monitor. If a motion sensor would first measure a reduction in forward progression (e.g. IMU close to the center of gravity like a phone in the back pocket or a sensor measuring the speed of the wheels of a walker), a heart rate monitor (e.g. smartwatch) could subsequently define whether this reduction in forward progression was voluntary (stop) when accompanied by a decrease in heart rate, or involuntary (FOG) when not accompanied by a decrease in heart rate. (IMU Inertial Measurement Unit; FOG freezing of gait)