Navigational cue effects in Alzheimer's disease and posterior cortical atrophy

Abstract

Objective: Deficits in spatial navigation are characteristic and disabling features of typical Alzheimer's disease (tAD) and posterior cortical atrophy (PCA). Visual cues have been proposed to mitigate such deficits; however, there is currently little empirical evidence for their use.

Methods: The effect of visual cues on visually guided navigation was assessed within a simplified real-world setting in individuals with tAD (n = 10), PCA (n = 8), and healthy controls (n = 12). In a repeated-measures design comprising 36 trials, participants walked to a visible target destination (an open door within a built environment), with or without the presence of an obstacle. Contrast and motion-based cues were evaluated; both aimed to facilitate performance by applying perceptual changes to target destinations without carrying explicit information. The primary outcome was completion time; secondary outcomes were measures of fixation position and walking path directness during consecutive task phases, determined using mobile eyetracking and motion capture methods.

Results: Results illustrate marked deficits in patients' navigational ability, with patient groups taking an estimated two to three times longer to reach target destinations than controls and exhibiting tortuous walking paths. There were no significant differences between tAD and PCA task performance. Overall, patients took less time to reach target destinations under cue conditions (contrast-cue: 11.8%; 95% CI: [2.5, 20.3]) and were more likely initially to fixate on targets.

Interpretation: The study evaluated navigation to destinations within a real-world environment. There is evidence that introducing perceptual changes to the environment may improve patients' navigational ability.

Figure 1

(A) Patient demographic information and background neuropsychological assessment; (B) Composite scores for visual processing domains and SRMT performance. The PCA group was more impaired on composite scores (Wilcoxon rank‐sum: Early: z = −2.32, P = 0.021; Visuoperceptual: z = −2.85, P = 0.004; Visuospatial: z = −2.05, P = 0.041), with weak evidence of greater impairment on the SRMT in the tAD group (z = −1.71, P = 0.088). Visual acuity but not contrast sensitivity was assessed; there was weak evidence for poorer acuity in the PCA than the tAD group (LogMar equivalent: z = 1.73, P = 0.085). Patients are arranged left to right in order of navigational performance from those taking the least to the most time to complete under the baseline (no cues) condition. Impaired scores (performance below 5th%ile) are highlighted in bold font. Mini‐Mental State Examination41; Short Recognition Memory Test42; Concrete synonyms43; Digit span forwards/backwards44; Graded difficulty arithmetic45; Graded difficulty spelling test46; Cortical Visual Screening Test47; Visual Object and Space Perception Battery48; Oblong edge ratio 1:1.2049; Letter Cancellation50; Usual/Unusual Views51; ^aUnpublished. *Healthy controls do not make errors. SRMT, short recognition memory test; PCA, posterior cortical atrophy; tAD, typical Alzheimer's disease.

Figure 2

(A) i – starting position, ii^1‐3 – obstacle positions (chosen to interrupt direct path to target destinations), iii^1‐3 – target positions, iv – participant position between trials; (B) point defined where participants reached target; (C) right door with obstacle under CCue + motion; arrows indicate direction of motion pattern movement. The setting was constructed at a pedestrian environment laboratory (PAMELA) able to simulate real world environments in a controlled manner.

Figure 3

Graph illustrating selected values of the odds ratio comparing secondary outcome measures between groups and cue conditions. For example, given an estimated odds ratio of 1.2, the figure shows that if the baseline FI or SI is 0.7 then FI or SI with the cue condition is expected to be 0.74, that is the cue condition increases the proportion of initial time spent fixating on the door (for FI) or shortens the path taken by the participant (for SI), compared with baseline. FI, fixation index; SI, straightness index.

Figure 4

Walking paths for control, tAD and PCA groups to left, middle and right door without obstacle, under baseline condition (no cues) generated using dead reckoning. Paths were estimated using foot velocity to calculate relative displacement between each step, and so do not show absolute position.27 Data are presented from when participants crossed the starting line. Coloured paths are particularly indirect relative to controls (<control mean SI – 3SD). First and last data points for walking paths are corrected to reflect trial start (x = 0, y = 0) and end positions (Left: x = 4.04, y = 2.4; Middle: x = 4.04, y = 0; Right: x = 4.04, y = 2.4). Circles represent end positions for censored trials. tAD, typical Alzheimer's disease; PCA, posterior cortical atrophy; SI, straightness index.