Objective Analysis of Performance of Activities of Daily Living in People With Central Field Loss

Abstract

Purpose: People with central visual field loss (CFL) adopt various strategies to complete activities of daily living (ADL). Using objective movement analysis, we compared how three ADLs were completed by people with CFL compared with age-matched, visually healthy individuals.

Methods: Fourteen participants with CFL (age 81 ± 10 years) and 10 age-matched, visually healthy (age 75 ± 5 years) participated. Three ADLs were assessed: pick up food from a plate, pour liquid from a bottle, and insert a key in a lock. Participants with CFL completed each ADL habitually (as they would in their home). Data were compared with visually healthy participants who were asked to complete the tasks as they would normally, but under specified experimental conditions. Movement kinematics were compared using three-dimension motion analysis (Vicon). Visual functions (distance and near acuities, contrast sensitivity, visual fields) were recorded.

Results: All CFL participants were able to complete each ADL. However, participants with CFL demonstrated significantly (P < 0.05) longer overall movement times, shorter minimum viewing distance, and, for two of the three ADL tasks, needed more online corrections in the latter part of the movement.

Conclusions: Results indicate that, despite the adoption of various habitual strategies, participants with CFL still do not perform common daily living tasks as efficiently as healthy subjects. Although indices suggesting feed-forward planning are similar, they made more movement corrections and increased time for the latter portion of the action, indicating a more cautious/uncertain approach. Various kinematic indices correlated significantly to visual function parameters including visual acuity and midperipheral visual field loss.

Vision has been shown to play an important part in how we plan and execute upper limb tasks such as reaching and grasping an object.^1–3 It is therefore hardly surprising that central field loss (CFL), due to AMD, has a significant impact on the ability to complete tasks requiring manual prehension.^4–7 

In order to be able to carry out everyday tasks, people with CFL usually adapt and employ some form of habitual strategy.⁸ People with CFL may, for example, use low-vision aids or adopt a nonoptical habitual strategy, which may change the manner in which a task is performed.⁹ Nonvisual habitual strategies include the use of other skills (e.g., exploiting tactile responses when filling a cup with liquid), moving closer to the task (a form of relative distance magnification), and optimizing lighting.¹⁰ 

The use of these adopted habitual strategies for completing activities of daily living (ADL) in people with CFL is somewhat unclear. Brody et al.¹¹ reported that a self-management program that educated people with AMD on adopting habitual strategies showed improved function among depressed patients only. Reeves et al.¹² reported no differences in patients' performance when comparing a ‘conventional' low vision rehabilitation program with an ‘enhanced' one, which included home visits from a rehabilitation officer. Lamoureux et al.¹³ found no improvement in the mobility and independence subscale of a questionnaire after rehabilitation training in a group of patients in which 62% were diagnosed with AMD. It is possible that these patients had already adopted a habitual strategy, which enabled them to complete the task sufficiently, resulting in no further improvement with training. The aforementioned studies^11–13 used ‘self-report' as the outcome measure. Previous research suggests that self-reported measures do not always accurately reflect patients' actual functional ability,^14–16 partly because this can also be influenced by other nonvisual factors such as depression.¹⁷ To improve upon the possible limitations of self-report, objective measures with qualitative assessments are employed.^18–21 One such objective method is three-dimensional (3D) motion analysis. This is regarded as the gold standard for measuring movement²² as it provides valuable additional detail of movement kinematics when compared with traditional measures that use basic measures such as ‘time to task completion.' 

Currently, there are very few studies that have used 3D motion analysis to investigate how these habitual strategies, adopted by people with CFL when carrying out ADL's, compared with those with healthy vision. Previous research that compared task performance between people with CFL and healthy vision used standardized procedures such that both groups completed the task in the same manner.^4–7 While this approach provided the opportunity to compare a standardized action between groups, it would have biased the results, as patients with CFL may have had to adopt an unnatural (nonhabitual) action to complete the task. This study investigates how the performance of individuals with CFL, when completing three everyday activities, compares with that of visually healthy subjects. Using objective 3D measurements, the study explored how participants with CFL carry out ADL's habitually, compared with age-matched visually healthy individuals. The three ADLs chosen were (1) identifying and picking up a carrot from a plate of food, (2) pouring liquid into a cup, and (3) inserting and turning a key in a lock. 

Methods

Participants

Fourteen people diagnosed with AMD (age 81 ± 10 years) and 10 age-matched visually healthy participants (age 75 ± 5 years) took part. There was no significant difference in the ages between the two groups (P > 0.05). Health and physical fitness of all participants was assessed through a self-report questionnaire, as used in previous studies.⁷ Participants were excluded if they reported any history of neurologic or musculoskeletal disorders (e.g., arthritis in upper body) that could affect manual prehension, vestibular disturbances, or failed to score the minimum requirement to pass the Mini Mental State Examination (MMSE).²³ 

Ocular diagnosis of bilateral macular problems was provided by a consultant ophthalmologist. Thirteen participants were diagnosed with AMD and one with macular hole. Participants were recruited from CamSight (a local charity based in Cambridge, UK, which supports people with visual impairment). The Tenets of the Declaration of Helsinki were observed and the experiment gained approval from the University's Ethical Committee. Written informed consent was obtained from each participant prior to undertaking the study. Among the participants recruited, five did not wear any refractive correction, four wore single vision distance spectacles, four wore single vision near spectacles, five wore bifocals, and six wore progressive powered lenses. It was assumed that the effect of wearing multifocals (compared with single vision spectacles) would have minimal effect on task performance since previous work from our laboratory has shown that near/intermediate manual prehension tasks are not affected by wearing multifocal when compared to single vision spectacles in healthy subjects (Lovell-Patel R, et al. IOVS 2013;54:ARVO E-Abstract 4250). 

Visual Assessments

Habitual binocular distance visual acuity (DVA) was measured with a retroilluminated Early Treatment Diabetic Retinopathy Study (ETDRS) chart at 4 m, using habitual distance vision spectacles. Near visual acuity (NVA) was measured with an externally illuminated near ETDRS chart at 40 cm using near spectacle prescription. Both distance and near acuity were measured on a per letter basis until no letters on a line could be correctly identified.²³ If participants were unable to read the largest letters at the initial distances, closer distances were used and the score adjusted accordingly. Contrast sensitivity (CS) was assessed binocularly using a Pelli-Robson chart at 1 m, with habitual distance prescription plus a +0.75 diopters (D) addition as per standard practice. Contrast sensitivity was measured per letter until no letters in a triplet could be correctly identified.^24–26 Table 1 gives individual and group characteristics. 

Table 1

 

Data for Visually Impaired (VI) and Visually Healthy Participants

 

Visual fields were measured with the Humphrey Field Analyzer (Carl Zeiss Meditec, Inc., Dublin, CA, USA). The central 24-2 SITA standard program was used with the participants wearing the required near vision correction. Monocular visual field results were integrated to produce a binocular field plot using the ‘best location' model.²⁷ To determine the extent of CFL among participants, the mean threshold^28,29 of the integrated field was calculated for the central 5°, central 10° and midperipheral 10° to 30° (Fig. 1). These identified the central 4 and 16 and outer 38 test points, respectively^7,29 (see Table 1 for individual and group visual fields). 

Figure 1

 

Integrated binocular visual field plot for a visually normal participant with the central 5°, 10°, and midperipheral 10° to 30° grids overlaid.

The Tasks/Procedures

Participants were required to complete the following ADLs; (1) identifying food on a plate and picking it up with a fork, (2) pouring liquid into a cup (up to a prespecified level), and (3) inserting a key in a lock and turning it. These tasks were chosen as they are common tasks, have a strong relevance to independent living and have been incorporated in several Visual Quality of Life (VQoL) instruments,^30–35 as well as in objective assessments of visual function.^20,21 

Participants with CFL were asked to complete each of the ADL tasks as they would do at home, wearing their habitual refractive correction, and using the hand they used to complete each of the tasks in their home. The participants were asked to identify which hand they used to complete each of the tasks at home. After providing sufficient time for the participants to familiarize themselves with each ADL, three trials were collected in each task. The familiarization process was only to make the participants aware of the individual elements of the task. So, for example, participants would pick up the bottle to see what it felt like, how heavy it was, its size, and so on. They did not carry out the task of pouring liquid from it. For the food on plate task, they picked up the fork to understand its weight. None of the participants reported that they would use any extra visual aids (e.g., magnifiers or torches) when carrying out these tasks at home. 

All subjects with healthy vision were asked to complete the tasks as they would normally, but were given additional specific instructions before data collection. For example, they were asked to use their dominant hand only (specific details for pertaining to each task are given below), make sure that the hand movement started from a specific starting position, and ensure that they did not change their working distance. A pilot study had confirmed how each of the tasks would be performed by a group of visually healthy participants. The subjects were also provided the opportunity to familiarize themselves with each ADL. All subjects reported that the specific instructions did not interfere with any of the tasks that they would carry out habitually. An average of three trials was obtained for each task. 

Identifying Food on a Plate Task.

Participants were seated in front of a table (76 × 76 cm) covered with a black cloth. A carrot stick, celery stick, and orange segment (each approximately 5 × 2 cm) were placed on a white plate (24.5-cm diameter) positioned 25 cm from the midline of their body (Fig. 2). The participant had to identify the carrot by reaching out and picking it up with a fork, and then lifting it from the plate. All participants could correctly identify the different food items on the plate. The starting position for each trial was the same for each participant: they were all required to hold the fork with their dominant hand which was placed 25 cm from the midline of their body lateral to the plate, at a premarked position (Fig. 2). Their nondominant hand was placed on the table on the other side of the plate 25 cm away from the midline of their body. Participants with CFL carried out the test as they would at home. 

Figure 2

 

Illustration of each ADL task: food on plate (A), pouring (B), and key in lock (C).

Pouring Task.

Participants were required to stand in front of a table 95 cm high, replicating the height of a typical kitchen work surface. The surface of the table was covered with a black cloth. Participants were asked to place their dominant hand at a predetermined position 25 cm from the participant's midline. A clear plastic bottle (height 32 cm, radius of widest part of the bottle 9.2 cm, volume 1.5 L) containing a dark liquid (0.75 L) was positioned 35 cm from the participant's dominant hand (Fig. 2). On hearing the ‘GO' command, participants were asked to pick up the bottle and pour the liquid into an empty white cup (8-cm width × 10-cm height) to within approximately 1 inch of the top. The cup was placed 25 cm from the bottle on the participant's midline. Participants with CFL were asked to carry out the task as they would at home (e.g., using both hands in the task and/or change viewing distance). 

Key in Lock Task.

Participants were required to reach out, insert a key into a Yale standard security night latch lock (Willenhall, West Midlands, UK), and turn the key until the lock released. The lock was built into a reduced section of a door. Participants were asked to place their dominant hand at a predetermined position 25 cm from the participant's midline. The lock was positioned 35 cm in front of the participant, in line with their hand (Figs. 2A–C). The lock was positioned 117 cm from the ground, reflecting the typical height of a lock found on a front door. Participants were required to hold the key with their dominant hand, insert it, and turn it. 

Kinematic data were collected (100 Hz) using a six camera 3D motion capture system (Vicon, 460; Oxford Metrics Ltd., Oxford, UK). Retroreflective spherical markers were attached at the following key anatomical points: distal border of the thumbnail and index fingernail, styloid process on the radial side of the wrist, sternum, and the anterio- and posteriolateral aspects of the head. 

Innovative approaches were used to precisely determine task completion/end of the movement for each ADL: (1) Food on plate: reflective tape was stuck to the plate under where the carrot was positioned. When the carrot was lifted from the plate, the Vicon cameras were able to track the reflective tape. The instant when the reflective tape appeared denoted the end of the movement, (2) Pouring: A reflective marker was attached to the lowest part of the bottle. The Vicon cameras were able to track the movement of the bottle, which subsequently made it possible to define when the bottle was repositioned on the table after the pouring task was completed, (3) Key in lock: reflective tape was attached to the Yale bolt. As the lock was turned, the reflective tape disappeared into the locking mechanism and the Vicon cameras were unable to track this marker. The instant when the reflective tape disappeared denoted the completion of the movement. 

Marker trajectory data were filtered using the cross-validatory quintic spline smoothing routine with ‘smoothing' options set at a predicted MSE value of 10 and processed using the PlugIn-Gait software (Oxford Metrics Ltd.). Relevant kinematic indices (Table 2) were measured for each of the tasks, which we and others have previously found important for upper limb movements.^4–7 

Table 2

 

Kinematic Variables Used to Analyze Each Activity of Daily Living, Including How Specific Aspects of the Movement Were Defined

 

Velocity Corrections.

Velocity corrections in the movement were calculated based on the work by Melmoth et al.³⁶ who quantified the extra movement or plateaus in the velocity profile after peak velocity. Figure 3A shows the typical forward velocity profile of an individual reaching to pick up the bottle (in the pouring task). In the early part of the reach, peak velocity is achieved (∼620 mm.s⁻¹) at approximately 0.06 seconds into the movement. This peak is followed by the individual reducing their velocity as their hand moves increasingly closer to the object (demonstrated by the velocity profile reducing back toward zero). During the latter part of the movement (∼0.13 seconds) there is an increase in velocity, which is demonstrated by a smaller (second) peak at approximately 0.17 seconds. To ensure the additional movements in the velocity profile are easier for the reader to see, we include Figure 3B, which plots the velocity trace in Figure 3A from 0.125 seconds into the reach. 

Figure 3

 

(A) Typical forward velocity profile of the hand reaching to pick up the bottle (in the pouring task) for patient with CFL. (B) More detail on the forward velocity profile of the hand shown in (A) from 0.125 seconds until end of movement.

Data Analysis

For each ADL, data were analyzed using a paired samples t-test (or nonparametric equivalent where relevant). Level of significance was accepted at P less than 0.05. For all dependent measures, variability was calculated as the SD across the three trial repetitions and analyzed using a paired samples t-test (or nonparametric equivalent where relevant). 

Results

Table 3 gives the mean and SDs for the different kinematic indices in participants with CFL and healthy subjects. Kinematic data show that the AMD participants made more movement corrections as well as increasing the overall movement time. 

Table 3

 

Mean and SDs for the Different Kinematic Indices for Participants With CFL and Healthy Subjects

 

Food on Plate Task

Participants with CFL had a significantly higher number of miss stabs (failed attempts to secure the carrot stick on the fork prior to lifting from the plate) compared with visually healthy individuals (P = 0.006). In participants with CFL, the number of failed attempts was 40. This averaged to three (±3) stabs per participant. Healthy subjects showed only one failed attempt in total. These trials were discounted. All participants were able to carry out the task successfully for three complete trials. 

Kinematic Analysis.

Overall movement time was significantly longer in participants with CFL compared with visually heathy participants (P = 0.005). Participants with CFL took twice as long to complete the movement. 

Velocities.

There were no significant differences in peak forward velocity (P = 0.347), or peak vertical velocity (P = 0.523) between the CFL group and healthy subjects; these kinematic indices indicate preplanning (feed-forward) part of the movement.² 

Vertical and forward velocity corrections, which occur during the latter part of the reach suggesting online control part of the movement³⁶ were 2.07 times and 2.62 times greater in CFL group compared with the visually healthy (P = 0.009 and P = 0.014, respectively). 

Deceleration Time.

This also gives an indication of the online control of the movement.^36,37 Vertical and forward deceleration times were 1.81 times and 2.13 times longer in the CFL group compared with the visually healthy (P = 0.007 and P = 0.009, respectively). 

Minimum viewing distance in the CFL group was significantly reduced to 79% of that of the healthy group (P = 0.043). 

Pouring Task

Participants with CFL typically used their nondominant hand to bring the cup closer to them. Participants also used a ‘double pour' method, whereby they would fill the cup to approximately three-quarters full, stop, and assess the volume of liquid in the cup, and then pour more liquid in to reach the desired level. Despite using this method, there were no significant difference in the number of spills between the CFL in the habitual condition and the visually healthy (P > 0.05). 

Kinematic Analysis.

The time from movement initiation to pick up the bottle was 45% longer in CFL participants compared with healthy participants (P = 0.037). 

Pouring time was significantly longer in visually impaired participants compared with healthy subjects (P = 0.031). It took CFL participants 26% longer to pour the liquid. 

Overall movement time was 34% longer in CFL participants compared with the visually healthy (P = 0.009). 

Velocities.

There was no significant difference in peak vertical and forward velocities of reach between the CFL group and the visually healthy (P = 0.829 and P = 0.947, respectively). These findings suggest that the planning or feed-forward aspect of the movement may not be significantly different between the two groups. Vertical velocity corrections during the latter part of the reach were also not significantly different in CFL participants compared with the visually healthy (P > 0.05). 

Deceleration.

Vertical deceleration time when reaching for the bottle was 42% slower in the CFL group compared with the visually healthy (P = 0.042). There was no significant difference in the forward deceleration time between the two groups (P = 0.085). 

Grip Aperture.

Maximum grip aperture was unaffected by group (P = 0.094). 

Minimum viewing distance in CFL participants was 86% of that used by healthy participants (P = 0.024). 

Key in Lock Task

Central visual field loss participants would use either their dominant or nondominant hand to feel for the specific point on the lock where the key should be inserted. 

Kinematic Analysis.

Overall movement time was nearly twice as long (1.94 times) in the CFL group compared with the visually healthy (P = 0.010). 

Velocities: Peak vertical velocity in CFL participants was 73% of that shown by healthy subjects (P = 0.002). Peak forward velocity was not significantly different between the two groups (P = 0.181). These findings suggest that, while the initial feed forward planning of the forward component of the reach is similar to normals, CFL participants planned a slower, more tentative vertical movement. 

Vertical and forward velocity corrections were 2.20 times and 2.27 times, respectively, more frequent in CFL compared with the visually healthy (P = 0.006 and P = 0.012, respectively) highlighting that they needed to make more frequent corrections during the latter (online) part of the movement. 

Deceleration: vertical and forward deceleration time were significantly longer in the CFL group compared with the visually healthy (P = 0.009 and P = 0.013, respectively). This may be because the CFL group initiated more online corrections in the latter part of the movement. 

Minimum viewing distance with CFL participants was 84% of that used by healthy participants (P = 0.029). 

Variability

Variability refers to the consistency of the movement over trial repeats. A larger variability (SD within the trials) indicates uncertainty in the movement planning/execution. Table 4 shows the mean variability of key kinematic indices in the two groups. As expected, the variability in participants with CFL was significantly greater. Interestingly, there was little difference in variability in peak forward and vertical velocities, and in the minimum viewing distance in the two groups, demonstrating more consistency for these kinematic indices. With other indices, more variability was found; for example, in the pouring task, a closer inspection of the movement profiles of participants picking up the bottle showed that they did not pick up the bottle in the same way from trial to trial. Sometimes the bottle was picked up in more of a forward motion (e.g., when the arm/hand moved toward the bottle, the bottle was picked up while the arm continued moving forward in an almost ‘sweeping' type motion). On other occasions, the hand slowed down such that the bottle was lifted vertically with minimal forward motion. This suggests that participants with CFL would have employed a number of different ‘adaptations or tactics' in order to make ‘online adjustments.' They may have responded differently in order to adapt quickly to unexpected/changing circumstances. As these movements are variable, they may well preclude ‘strategy forming' responses which need to be constant in order for the movement to habitual. 

Table 4

 

Mean Variability in CFL Participants and Healthy Visuals

 

Correlation With Visual Function

To understand how the level of participants' visual impairment affected the execution of the task, correlation analysis was performed between the level of vision (DVA, NVA, CS, visual fields) and the various kinematic indices that showed a significant difference. These analyses used data from CFL participants only. 

Distance acuity was significantly (P < 0.05) correlated to near acuity (r = −0.637), contrast sensitivity (r = −0.545), central 5° field (r = −0.545), central 10° field (r = −0.681), and midperipheral 10° to 30° field (r = −0.565). This is well reported in the literature. 

For the food on plate task, there were no significant associations with any of the visual function parameters. For the pouring task, the forward reach velocity correlated with 10° to 30° visual field (r = −0.629, P = 0.016), overall movement time correlated with 10° to 30 ° visual field (r = −0.560, P = 0.037), and the pour time (from grasping bottle to placing back on table after pouring) correlated with 10° to 30 ° visual field (r = −0.627, P = 0.016) and CS (r = −0.586, P = 0.028). All the associations were negative. This indicates that the pouring task was significantly influenced by the midperipheral visual field loss. The pour time was also influenced by contrast sensitivity. For the key in the lock task, the vertical deceleration time correlated with central 5° visual field (r = −0.657, P = 0.015), central 10° visual field (r = −0.599, P = 0.030), the minimum viewing distance correlated with DVA (r = −.715, P = 0.006), NVA (r = −0.713, P = 0.006), and central 10° visual field (r = 0.580, P = 0.038). This indicates that the key in lock task was significantly influenced by the central visual field loss and acuity levels. 

Discussion

More than 90% of people with visual impairment report using some form of habitual strategy to aid function in ADL.⁸ This is the first study to use detailed kinematic analysis to compare the performance of people with CFL, using habitual strategies, with visually healthy individuals. We used the accepted gold standard method of objective analysis by using 3D motion analysis (Vicon) to investigate whether participants with CFL completed three everyday tasks as well as visually healthy subjects. 

Various different adaptations were observed in patients with CFL. These included a shorter working distance (either leaning closer or moving the object closer to them) and the use of both hands. Patients with CFL made more movement corrections as well needing an increased overall movement time. In addition, kinematic data show various other differences for the different tasks. 

Food on Plate Task

Despite the adoption of habitual strategies, participants with CFL took longer to complete the task (Table 3). Kinematic indices suggesting the initial planning (feed-forward) aspect of the movement in participants with CFL were similar to the visually healthy. It has been suggested that peak velocity is mediated through feed-forward mechanisms, representing the planning component prior to movement onset.^2,36 In our study, neither vertical and forward peak reaching velocities were significantly different between groups, indicating that participants with CFL performed just as well as the visually healthy (Table 3). 

Participants, however, required longer times during the latter (online control part) of the movement to initiate corrections and ‘fine tune' the movement, as shown by significant differences in both forward and vertical deceleration times and number of velocity corrections (Table 3). 

Pouring Task

Participants with CFL were slower in the time taken to reach out and pick up the bottle and complete the pouring aspect of the task. This subsequently resulted in a longer overall movement time. Central visual field loss participants planned the execution of the movement in a similar manner to healthy subjects, as suggested by similar peak vertical and forward velocities, when reaching for the bottle. They did, however, exhibit longer vertical and forward deceleration times (although only vertical deceleration time was significant) compared with healthy subjects. 

Key in the Lock Task

Participants with CFL took significantly longer to complete the task compared with the visually healthy. This is due to a combination of participants planning a slower initial vertical part of the reach, and longer vertical and forward deceleration time (Table 3). These findings have been similarly reported by Timberlake et al.⁶ with AMD and visually healthy participants who were required to reach and grasp an object positioned directly in front of them. 

Participants with CFL showed greater uncertainty evidenced by a higher number of forward and vertical velocity corrections. Online corrections were higher resulting in significantly increased forward and vertical deceleration times. In addition, the minimum viewing distance was significantly shorter in participants with CFL compared with healthy subjects. 

Taking all the tasks together, important similarities and differences exist between the two groups and the tasks used in the current study. Participants with CFL took longer to complete the tasks despite the fact that the minimum viewing distance was significantly reduced, which would, in turn, have resulted in an improved visual resolution of the object. The vertical deceleration time was also significantly different to healthy individuals, indicating longer time for online control (or latter part of the movement). Only in the key in the lock task was the feed-forward aspect, suggested by vertical velocity, slower when compared with the other tasks in participants with CFL. This may be due to the ‘difficulty' of the task compared with the others. Both food on the plate and the key in the lock task showed increased numbers of online corrections compared with the pouring task. This suggests that the habitual strategies for these were not as effective as the pouring task. The improvement in performance of the pouring task may be explained by the fact that participants with CFL used a ‘double pour' method (resulting in an increased overall movement time), which may have negated the need for making other smaller online corrections. 

Previous research has assessed self-reported difficulty of various ADLs in visually impaired observers, using item response theory to determine item difficulties for a number of different tasks.¹⁷ Higher (more positive) item difficulty indicates that respondent must have a higher level of functional ability to be able to complete the task.³⁵ The item difficulty for pouring (+0.03 logits) indicates that it is perceived to be a harder task to complete than inserting a key in a lock (−0.09 logits) or seeing food on a plate (−0.20 logits).¹⁷ The three tasks examined objectively in our study showed significantly lower performance when compared with healthy subjects. In order to rate the difficulty of the task, we compare the percentage difference in time to complete each task between groups. This measure has been used in previous research to establish ‘difficulty' in carrying out tasks.^15,18–21 Central visual field loss participants took twice as long (3.60 vs. 1.76 seconds) for the food on the plate task, twice as long for the ley in lock task (5.86 vs. 3.01 seconds), and 45% longer (1.63 vs. 1.12) for the pouring task when compared with healthy participants. In addition, CFL patients also made more velocity corrections in the food on plate and key in lock tasks compared with pouring task. Therefore, while self-report research indicates that the pouring task is the hardest while the key in lock is the easiest, objective measures in this study suggest that the pouring task to be the easiest while the key in the lock and food on the plate are equally difficult. There may well be differences between the two assessment measures as self-report considers ‘what you do in the home,' whereas our tasks were measured in the laboratory. In addition, the fact that participants were not required to pour the liquid to a specified level accurately may have contributed to this effect. We would expect that the pouring task to be harder had this been a condition of the experiment. We would then also expect them to be slower. Nevertheless, important differences exist between self-report data and our objective measures. 

It may be assumed that the closer viewing distance employed by the CFL subjects also ‘improved the visual resolution of the object.' This is undoubtedly true and is possibly the reason as to why the subjects chose to lean forward. However, this does not change our conclusions of how they completed the tasks because the starting position of the hand was not changed. The hand had to travel the same distance for all participants. 

It is possible that indices that have been associated with ‘online control' may be influenced by ‘feed-forward' mechanisms. However, the kinematic indices associated with feed-forward planning were similar in the two groups so this would not have contributed to the differences in online control. It is also expected that the feed-forward mechanism is sufficient to enable the hand to be placed in the approximate area of the object/end point, and therefore differences in kinematic indices, which represent the latter part of the action, suggest that the online control was affected. In addition, variability, which refers to the consistency of the movement over trial repeats or the certainty of that aspect of the task being completed, show no differences between the two groups in peak velocity but they do in a number of other online variables. 

Our results also show significant correlations between some visual function scores and task performance. For example, the key in the lock task was associated with visual acuity demonstrating that the task required good resolution. Various other kinematic indices correlated significantly with 10° to 30° visual field, suggesting that this parameter plays an important part in enabling the completion of these tasks. It is likely that midperipheral visual field enables important information, such as the position of the hand and the wrist, to be used hence exhibiting its importance. Previous research, including from our laboratory, has shown that when peripheral visual field loss of up to 11° is simulated using a pinhole, then indices which suggest both planning and online control of a simple reach and grasp task are decreased.³⁸ Midperipheral visual field has also been shown to be important in explaining self-reported difficulty in mobility tasks^17,29 and visual information tasks.¹⁷ Interestingly, contrast sensitivity did not correlate significantly to many kinematic indices of the tasks of this study. This was surprising because it has been shown to correlate significantly to indices representing straightforward reaching and grasping tasks.^4,5 Only the pouring task in the present study showed a significant correlation with CS. Tabrett and Latham,¹⁷ from self-reported data, reported that although CS was significant visual function predictor for reading and mobility tasks, it did not correlate significantly to visual motor tasks (including the tasks ‘pour or mix without spilling' and ‘insert key into lock') or visual information tasks (including ‘seeing food on a plate'). It is therefore likely that the importance of contrast sensitivity in completion of tasks may be dependent on the type of motor task and its complexity, as well as the experimental conditions including lighting. The pouring task (with a clear jug) would have posed a low contrast task, which needed more input from the participants' contrast sensitivity than other tasks which were of higher contrast. This clearly needs further investigation. 

Conclusions

Results indicate that, despite the adoption of various habitual strategies, participants with CFL still do not perform common daily living tasks as well as visually healthy subjects. Although the kinematic indices, which suggest feed-forward planning, were generally similar, a greater number of online corrections were required and the time taken during the online control period was also longer in CFL participants. These results indicate that further rehabilitation may be beneficial. On the other hand, it may be the case that the variability of movements shown during online control in people with CFL may mean that they reach a threshold beyond, which no further performance improvement would be possible; further research would investigate this. 

Acknowledgments

The authors thank Kieran Turner and Amy Scarfe, PhD, for their support with data processing and data analysis.  

Disclosure: S. Pardhan, None; K. Latham, None; D. Tabrett, None; M.A. Timmis, None