Visual blur disrupts the kinematic and temporal aspects of reach-grasp-lift movements

1 1 Visual blur disrupts the kinematic and temporal aspects of reach- 2 grasp-lift movements. 3 William E. A. Sheppard 1,2, Carlo Campagnoli1, Richard M. Wilkie1, Rigmor C. Baraas3, and Rachel. 4 O. Coats 1 5 Affiliations: 1 School of Psychology, Faculty of Medicine and Health, University of Leeds; 2 Sheffield 6 Centre for Health and Related Research, School of Medicine and Population Health, Faculty of 7 Health, The University of Sheffield; 3 National Centre for Optics, Vision and Eye Care, Faculty of 8 Health and Social Sciences, University College of Southeast Norway, Kongsberg, Norway 9 Corresponding Author: William Sheppard 10 Corresponding Author email: [email protected] .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 2 11 Abstract 12 Degraded vision (caused by pathological reasons or monocular viewing) has been shown to affect fine 13 motor control. However, there is a dearth of work examining the effects of “cataract-like” blur on 14 reach-to-grasp performance. There is, however, a trend towards amblyopic blur being associated with 15 deficits in reach-to-grasp performance, suggesting that timely intervention in treating cataracts is 16 likely to be essential to maintain a functional ageing population. 18 participants performed a reach-to- 17 grasp task. They reached for and precision grasped high and low-contrast cuboid targets under three 18 visual conditions: binocular blur, monocular blur (full vision in the other eye) and full vision. They 19 also performed contrast sensitivity, stereoacuity and visual acuity tests. Visual blur was associated 20 with changes to the kinematics of prehensile movements' early/acceleration stage (maximum 21 acceleration and maximum velocity) and maximum grip aperture. Visual blur also caused the period 22 from first contact with the target to the time it was lifted (dwell time) to be elongated. These results 23 suggest that changes in prehension associated with visual blur are linked to differences in the planning 24 and online control of prehension movements. 25 26 27 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 3 28 Introduction 29 The ability to move one's arm and hand towards an object to grasp it (prehension) is an essential 30 component of many everyday actions, including reaching for a door handle or picking up a mug of 31 tea. Therefore, factors affecting the execution of these movements can directly and significantly 32 impact one's quality of life. The spatiotemporal characteristics of prehensile movements depend on 33 three fundamental aspects: the visual system's internal state, the target object's visual characteristics 34 and the interaction between the two (1). To complete a prehensile movement, the individual must use 35 their representation of these three aspects to estimate the hand and target positions (2) before planning 36 the path between the two, as well as the final orientation of the grip (3). Once an individual has 37 initiated the movement, they must effectively control it to its completion, which ideally requires 38 visual and proprioceptive feedback. 39 Given the central role of vision in planning and executing prehensile movements, it is essential to 40 understand how degraded vision affects reaching and grasping actions. There has been extensive 41 research demonstrating the impact of monocular occlusion on both the planning and execution (online 42 control) of prehensile movements (4–9) and a significant number of studies investigating prehension 43 under monocular and/or binocular visual conditions, such as in the case of amblyopia (1,10–16), age- 44 related macular degeneration (AMD (17–21)) and glaucoma (17,20,22–24). 45 Grant and Conway (2019) documented the effect of monocular occlusion on the spatial and temporal 46 aspects of prehension. Specifically, monocular occlusion was associated with decreased maximum 47 velocity (MA), increased maximum grip aperture (MGA) and longer movement time, primarily 48 caused by an extended deceleration phase and contact-to-lift duration (dwell time). Interestingly, the 49 effect of monocular vision on MGA and dwell time disappeared when visual feedback during the 50 movement was removed, suggesting that monocular vision directly impacts the online control of the 51 movement. In contrast, the aspects of the movement related to planning (time to MV and the time to 52 MGA) were not affected by the removal of visual feedback. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 4 53 Similar research has been conducted in anisometropic amblyopia patients, although the findings were 54 more heterogeneous. There is consensus that amblyopia is associated with an increase in total 55 movement time (10,11,14) and dwell time (10,11), as well as a trend towards a decrease in MV 56 (11,14). However, as shown in Table 1, other variables, including maximum acceleration (MA), 57 duration of the acceleration phase, MV, duration of the deceleration phase, and MGA, do not show a 58 consistent pattern or are inconsistently reported (10,11,14). 59 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 5 60 Table 1. A summary of the effects of amblyopia on kinematic markers in prehensile tasks. The effects 61 in the table are those associated with amblyopia vs. controls, i.e. “Increased” suggests that amblyopia 62 was associated with an increase in a variable relative to controls. Buckley et al. (2015) Grant et al. (2007) Niechwiej- Szwedo et al. (2011) Movement time Increased Increased Increased Maximum velocity No significant effect Decreased Decreased Maximum acceleration Not reported Not reported Decreased Duration of the acceleration phase No significant effect Not reported Decreased Maximum deceleration Not reported Not reported No significant effect Duration of the deceleration phase Not reported Not reported No significant effect MGA Decreased No significant effect Not reported Dwell time Increased Increased Not reported 63 64 Reduced contrast between stimuli and background also appears to magnify amblyopia's effects on 65 prehension. For example, amblyopes showed a larger increase, relative to controls, in planning time 66 and grip aperture at contact with low contrast targets relative to the high contrast targets (1). The 67 performance deficits demonstrated in those with amblyopia did not correlate with reduced 68 binocularity or VA, but reduced CS was associated with increased target localisation errors (1). This 69 is not surprising as amblyopia and visual blur have been demonstrated to degrade CS by increasing .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 6 70 spatial sensitivity thresholds across the spectrum of spatial frequencies (25–27), with the effect being 71 particularly strong at higher spatial frequencies associated with fine depth perception (28). 72 At the point of writing, there is little to no research on the effects of cataracts and cataract removal 73 surgery on prehension. According to The Royal College of Ophthalmologists 2022 report, over 74 450,000 cataract removal procedures were carried out in England in the financial year 2019/20, an 75 11% increase from three years earlier. To complicate matters, the medical history of cataract patients 76 is highly variable and heterogeneous, raising concerns about the generalizability of the findings. 77 Therefore, in the present study, we aimed to systematically assess the effects of cataract-like visual 78 blur on grasping, which, at the time of writing, has not been directly assessed in previous research. 79 We manipulated healthy participants' vision by artificially blurring one or both eyes' visual field, since 80 cataracts can develop in either eye (monocularly) or both eyes (binocularly), whilst they performed a 81 reach-to-grasp task to high and low contrast targets. We predicted that increased blur would be 82 associated with changes in the movement's spatial and temporal aspects, particularly in the scaling of 83 MGA, the duration of overall MT, the duration of the deceleration phase, and the duration of dwell 84 time. Additionally, the effects of degraded vision would likely be magnified when the participants 85 reached for low-contrast targets. These findings have the potential to contribute to the growing body 86 of evidence for the importance of timely cataract removal. 87 Methods 88 Participants 89 An opportunity sample of 19 participants completed the study between 24/10/2023 and 07/11/2023. 90 One participant repeatedly failed to follow the task instructions, so the analysis did not include their 91 data. 92 The ages of the final sample of 18 participants (11 female, 7 male) ranged from 20 to 31 (mean = 93 24.23 years, sd = 3.36 years) years old. All participants reported having normal or corrected-to- 94 normal vision, and all participants reported being right-handed. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 7 95 Participants were excluded if they were under 18, had persistent low vision that could not be corrected 96 with glasses, had any disorder/condition that may affect their ability to grasp objects/impair their 97 coordination, or were left-handed. Participants gave written consent for their participation in the 98 study. Upon completing the study, the participants were entered into a prize draw to win a £30 99 Amazon voucher. The University of Leeds School of Psychology Ethics Committee granted ethical 100 approval on 12/05/2023 (Ethics Reference Number: PSYC-899). 101 Design 102 The present study employed a within-subjects experimental design whereby participants completed 103 four tasks (reach-grasp-lift [RGL], contrast sensitivity [CS], visual acuity [VA] and stereoacuity) 104 under three visual conditions (full vision, monocular blur and binocular blur). Participants completed 105 the RGL, CS, and VA tasks in the first testing session in that order, with the order of the visual 106 conditions being counterbalanced. Due to an oversight by the researchers, stereoacuity data were not 107 collected during the initial testing session and were obtained at a later testing session. 108 Within each visual condition, the RGL task was varied across two levels of contrast (‘high’, a wooden 109 target against a black background; ‘low’, a wooden target against a wooden background) and three 110 distances (150 mm, 250 mm, 350 mm), giving rise to a 3 (visual condition) x 3 (distance) x 2 111 (contrast) within-subject design. The participants repeated each condition eight times, giving 144 112 trials split into three blocks of 48 trials by visual condition. The experiment lasted 60 to 90 minutes in 113 total. 114 Procedure 115 Upon arrival, participants were presented with an information sheet outlining what would be involved 116 in the study and allowed to ask the researcher any questions. The participants read and signed a 117 consent form confirming their eligibility and willingness to participate in the study. Finally, 118 participants completed the eye dominance test. The researcher then fitted the participants with the first 119 pair of glasses (clear lenses over both eyes, a clear lens over the dominant eye and a blurred one over 120 the other, or blurred lenses over both eyes), depending on the visual condition performed first. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 8 121 Each participant sat as close to the table as possible with their navel in line with the midline of the 122 table. They started each trial gripping the start point: a round wooden dowel at the near edge of the 123 table, between their thumb and index finger, making sure all markers were visible to the cameras. 124 Upon a go-signal, the participants reached for the target as quickly and accurately as possible, 125 gripping the target dowels with their thumb and index finger before lifting it. The participant then held 126 the target suspended in the air. The researcher stopped the recording before the participant replaced 127 the target object on the table, and the trial ended. The participants repeated this procedure until they 128 completed all the trials for that visual condition and were then taken to a separate room to complete 129 the CS and VA tasks. This procedure was repeated for the three visual conditions before the 130 participant was debriefed and allowed to leave. 131 In the second testing session, the participants completed the stereoacuity task once for each visual 132 condition, in the same order as they had completed the initial testing session. The participant was then 133 debriefed and allowed to leave. 134 Materials and stimuli 135 The table used for the experiment was 1200 mm wide x 800 mm deep x 730 mm high and was painted 136 black. A board was mounted on the table that was 608 mm wide x 826 mm deep x 8 mm high. The 137 board was slid in the Y-plane to change the contrast condition without the participant moving. 138 Runners positioned at 456 mm from the midline of the table limited the board's movement. The left 139 half of the board was plain wood, and the right half was painted black. These created the low and high 140 contrast conditions, respectively (the target was wooden). The layout of each half of the board was 141 identical so that the midline of each half of the board was populated with a round wooden dowel of 27 142 mm diameter on the near edge (the starting point for each trial) and three raised rectangular grooves 143 (target locations) at 177 mm, 277 mm and 377 mm from the near edge of the table (150 mm, 250 mm 144 and 350 mm from the far edge of the starting point). See Figure 1 A for a photo of the experimental 145 setup and Figure 1 B for a schematic of each half of the board. 146 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 9 147 Figure 1. A, photo of the experimental setup used in the present study. B, a schematic of the 148 experimental setup used in the present study. 149 [INSERT FIGURE 1 HERE] 150 Lightweight infrared reflective markers (12.7 mm diameters) were attached to the index fingernail, the 151 index knuckle, the thumbnail and the wrist (head of the radius). The instantaneous position of the four 152 markers and one placed on the top of the target were recorded at 120 Hz with submillimetre spatial 153 resolution by four motion capture cameras (Optitrack Flex 13 [NaturalPoint, Corvallis, OR, USA]) 154 The target objects were cuboids (29 mm wide x 36 mm depth x 86 mm height) with dowels fitted to 155 two opposite sides (7 mm length x 10 mm diameter), as shown in Figure 2. These were positioned on 156 the board so that the front edge subtended an angle of 31 degrees to the front of the board. During a 157 pilot study, this angle was determined to give participants a comfortable reach. 158 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 10 159 Figure 2. Target schematic from the front (left) and top (right) elevations. 160 [INSERT FIGURE 2 HERE] 161 Contrast sensitivity (CS) was measured using a right and left-eye Pelli-Robson CS chart (Pelli et al., 162 1988) calibrated at a 1m distance. Visual acuity (VA) was measured using the Bailey Lovie Chart 4, 163 calibrated at a 6 m distance (30). Stereoacuity was measured at 40 cm using the TNO stereoacuity test 164 (31,32). Eye dominance was determined using an 'alignment test' (33). 165 Statistical analysis 166 The effect of the visual conditions on CS, stereoacuity, and VA was estimated using a standardised 167 procedure after calculating the thresholds. A multilevel approach (using a Generalised Linear Mixed 168 Model; GLMM) was taken (34). For each vision test score, the visual condition was entered into the 169 model as a fixed factor, with a random intercept within each participant to model individual 170 differences in each visual measure. 171 A similar GLMM approach was taken for the RGL data. Each outcome measure (listed and defined 172 below in Table 2) was modelled as a function of three fixed factors: visual condition, contrast and 173 distance. The maximal model included the three main effects plus all the two-way and the three-way 174 interactions and covariates (the CS, VA, stereoacuity and session number). If a fixed effect was not 175 significant and not included in the minimal model (the main effects of visual condition, contrast, 176 distance, covariates and the two-way interaction of visual condition and contrast), it was removed 177 from the model in order to achieve a parsimonious model; these are represented by a “–” symbol in 178 the model tables. Some non-significant predictors excluded from the minimal model can still be seen 179 in Tables 3, 4 and 5; this is due to each interaction effect containing several comparisons; for 180 example, if we look at the two-way interaction of contrast and distance, this includes two 181 comparisons: contrast x near vs. mid distance and contrast x near vs. far distance. If contrast x near vs. 182 mid distance produces a significant effect and contrast x near vs. far distance does not, both 183 comparisons will remain in the final model table, with the non-significant results represented as NS. 184 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 11 185 Table 2. The definitions of the kinematic and timing measures and their abbreviations. Parameter (abbreviation) Description (units) Movement planning Maximum acceleration (MA) Maximum wrist acceleration before contact (mm/s2). Maximum velocity (MV) Maximum wrist velocity before contact (mm/s). Maximum grip aperture (MGA) The maximum distance between the thumb and the index finger markers before contact (mm). Path length (PL) The cumulative distance travelled by a given finger or marker (mm). Time from onset (O) to MA (OtoMA) The time from movement onset to maximum acceleration as a proportion of MT (%). Time from MA to MV (MAtoMV) The time from maximum acceleration to maximum velocity as a proportion of MT (%). Time from O to MGA (OtoMGA) The time from movement onset to maximum acceleration as a proportion of MGA (%). Online control Maximum deceleration (MD) Maximum wrist deceleration before contact (mm/s2). Time from MV to MD (MVtoMD) The time from maximum velocity to maximum deceleration as a proportion of MT (%). Time from MD to first contact (MDtoC) The time from maximum deceleration to first contact as a proportion of MT (%). Dwell time (DwT) The time from first contact to the object lift as a proportion of MT (%). Movement time Movement time (MT) The time from movement onset to first contact (s). 186 Note: Movement onset was defined as the time point when the wrist velocity first exceeds 0.05m/s in 187 the Z plane. First contact was defined as the first time point when the index finger was behind the .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 12 188 target, and the velocity was less than 0.05m/s in the Z plane; if, once behind the target, the index 189 velocity never fell below 0.05m/s, the time point of minimum velocity was used. The lift time was 190 defined as the first time point at which the target exceeded 0.05m/s in the Y plane. These data were 191 extracted using the kinesis package (35) in R Statistical Software, version 4.3.1 (36). 192 The dependent variables were grouped into three categories to systematically examine how visual blur 193 influences reach-to-grasp movements: movement planning, online control, and movement time. This 194 classification reflects distinct phases of the prehensile movement, each associated with different 195 aspects of motor execution and sensorimotor feedback. Movement planning variables primarily 196 capture feedforward control mechanisms based on visual input before movement execution. They are 197 labelled as follows: maximum acceleration [MA], maximum velocity [MV], maximum grip aperture 198 [MGA], path length [PL], time from onset [O] to MA [OtoMA], time from MA to MV [MAtoMV], 199 and time from O to MGA [OtoMGA]. Online control variables reflect adjustments made during 200 movement execution. They are labelled as follows: maximum deceleration [MD], time from MV to 201 MD [MVtoMD], time from MD to first contact [MDtoC], and dwell time [DwT]. Movement Time 202 [MT] represents the overall temporal duration of the action and integrates aspects of movement 203 planning and online control. 204 A random intercept was estimated for each participant to account for differences in vision and 205 coordination between participants, and random slopes for distance and session number to account for 206 variability in participant arm length and learning rates, respectively. 207 Regarding the time series analysis, six kinematic landmarks were selected to divide each reaching 208 action: maximum acceleration (MA), maximum velocity (MV), maximum grip aperture (MGA), 209 maximum deceleration (MD), first contact with the target (Contact) and when the participant lifts the 210 target (Lift). These landmarks are shown in Figure 3 on an exemplar velocity curve. 211 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 13 212 Figure 3. An example velocity profile showing the landmarks used to break down each movement. 213 [INSERT FIGURE 3 HERE] 214 Note: The analyses will consider the time from the onset of movement to maximum acceleration 215 (MAcc), the time from the onset to maximum grip aperture (MGA), the time from maximum 216 acceleration to maximum velocity (MVel), the time from maximum velocity to maximum 217 deceleration (MDec), the time from maximum velocity to contact (Contact) and the time from contact 218 to lift (Lift, also known as dwell time) as a percentage of total movement time for that reaching 219 action. 220 The normally distributed outcome variables with no time component were modelled using a Gaussian 221 distribution (MA, MV, and MD). In contrast, the remaining variables were modelled using the 222 Gamma distribution because they have a zero-bound time component and are typically positively 223 skewed, which aligns with the distribution's assumptions. These variables were initially also modelled 224 using the Inverse Gaussian distribution. However, after the performance of each model was assessed 225 by comparing the BIC (Bayesian Information Criterion), all models were fitted using the Gamma 226 distribution. A BIC difference greater than 10 gives "very strong" evidence that the model with the 227 lower BIC value best fits the data, favouring models with lower numbers of factors (37,38). As all 228 outcome variables were scaled to improve model fit, a small constant was added to each value 229 (modelled with the Gamma or Inverse-Gaussian distributions), which only shifts the intercept and 230 does not affect regression slopes (i.e. β coefficients) or significance tests. These analysis scripts are 231 stored in a GitHub repository [https://github.com/willsheppard9895/blurNprehension], and the data is 232 stored in the figshare repository [https://figshare.com/articles/dataset/Prehension_Data/28877057]. 233 The outcome variables and numeric predictors (CS, stereoacuity, VA and session number) were 234 centred and normalised. As the Inverse Gaussian and Gamma distributions only accept positive 235 values, the outcome measures modelled using these distributions were further transformed to be 236 positive (by adding the lowest integer to create a set of positive values; this was calculated for each 237 variable), as per equation 1. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 14 x.scaled = x ― mean(x) sd(x) + minimum integer [1] 238 Visual condition, contrast, and distance were entered into each model as categorical predictors. 239 Regarding visual condition, the participant's performance was compared between binocular and 240 monocular blur and between monocular blur and full vision to match the vision changes associated 241 with first and second eye cataract surgery, respectively. Regarding distance, comparisons were made 242 between near and mid distance and near and far distance. Significant three-way interactions are 243 reported and plotted in the results section. Post-hoc tests were not conducted for significant three-way 244 interactions due to the large number of pairwise comparisons (N = 153), which would greatly increase 245 the likelihood of type I errors. Regarding significant two-way interactions, only those including visual 246 condition will be reported and plotted in the results section. Any significant visual condition x 247 contrast interactions will also be subject to post-hoc testing, as these interactions addressed the main 248 aims of the study. Significant main effects of visual condition and contrast are also reported in the 249 text. All MLM results are reported in Tables 3 (vision tests), 4 (movement planning), 5 (online 250 control) and 6 (movement time). If one or more significant three-way interactions predict an outcome 251 variable, the results section will not discuss any significant two-way interactions. 252 Inter-individual variation in each outcome measure was also assessed. For each outcome measure, the 253 heterogeneity within the sample was assessed by comparing the relative size of the SD of the random 254 intercepts allocated to each participant to the fixed intercept. As many of the intercepts were altered 255 by the transformations applied to make them positive, this transformation was removed for this 256 analysis. In the present case, this took the form σ β , where σ is equal to the magnitude of the random 257 intercept SD , and β is equal to the magnitude of the fixed intercept minus the transformation. When 258 this value exceeds 0.25, we concluded that the data are heterogeneous, as a participant at the 2.5th 259 percentile would have a score equivalent to 0.5 of the mean, and a participant at the 97.5th percentile 260 would have a score 1.5 times the mean (39). These results were only reported if the effect was 261 heterogeneous. The random effects of session number and distance were not subject to this analysis, .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 15 262 as these were entered to account for inter-individual variability in learning rate and arm length, 263 respectively, and did not speak to the overall aims of the manuscript. 264 All analyses were performed using R Statistical Software (36). Kinematic measures were extracted 265 from the RGB data using the Kinesis package (35), GLMMs were estimated using the lme4 package 266 (40), p-values were estimated using Satterthwaite's approximation through the lmerTest package (41), 267 and estimated marginal means (EMM) for post-hoc testing were estimated using the emmeans 268 package (42). 269 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 16 270 Results 271 Vision tests 272 To ensure that blurred conditions were affecting vision in a way similar to cataracts, an MLM analysis 273 was carried out on the effects of visual conditions on contrast sensitivity (CS), stereoacuity, and visual 274 acuity (VA) tasks. The β values in Table 3 predict the mean CS, stereoacuity and VA. 275 Table 3. MLM estimates for fixed and random effects predicting performance on contrast sensitivity, 276 stereoacuity and visual acuity tasks. Estimate Contrast sensitivity Stereoacuity Visual acuity Intercept 1.80*** [1.77, 1.83] 1.01*** [0.69, 1.34] 1.07*** [0.94, 1.19] Fixed effects (β) Binocular blur vs. monocular blur 0.18*** [0.14, 0.21] −0.57** [−0.91, −0.24] −0.28*** [−0.31, −0.25] Monocular blur vs. full vision 0.12*** [0.08, 0.15] −0.39** [−0.62, −0.16] −0.18*** [−0.21, −0.15] Random effects (σ) Contrast sensitivity Stereoacuity Visual acuity Participant ID 0.04 0.43 0.12 Model fit R2 marginal 0.668 0.077 0.366 R2 conditional 0.872 0.399 0.859 Note: The participant's performance on the stereoacuity tasks was standardised (i.e., with a mean equal to zero and SD equal to one). None of the outcome variables were normally distributed and were, therefore, modelled using the Gamma distribution. The Gamma distribution only accepts positive variables; thus, stereoacuity and visual acuity were made positive by adding 1. Due to the stereoacuity scores being centred, the β values for the stereoacuity task are expressed as a proportion of the grand SD of stereoacuity (66.14 arcsecs). NS = p ≥ .05, ** = p < .01, *** = p < .001. The numbers in each cell represent the estimate [lower 95% CI, upper 95% CI]. 277 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 17 278 MLM analysis found significant effects of visual condition on CS (see Figure 4A), stereoacuity (see 279 Figure 4B) and VA (see Figure 4C); monocular blur (relative to binocular blur) and full vision 280 (relative to monocular blur) improved CS by 0.18 and 0.12 log units, stereoacuity by 37.93 and 25.79 281 arc secs and VA by 0.28 and 0.18 logMAR. 282 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 18 283 Figure 4. Effects of visual conditions on each vision test. The central black line of each violin 284 represents the mean, and the outer lines represent the range. A. CS (log units). B. Stereoacuity (arc 285 secs). C. VA (logMAR). 286 [INSERT FIGURE 4 HERE] 287 The SD of the random intercepts was 4300.00% and 171.43% of the fixed intercepts for stereoacuity 288 and VA, respectively. This indicates extremely high individual differences in visual functions and that 289 group-level averages may obscure meaningful variability influencing prehensile performance. 290 Movement planning 291 The β values in Table 4 predict maximum acceleration, µ ma, maximum velocity, µmv, maximum grip 292 aperture (MGA), µ mga, path length, µpl, the time from onset to maximum acceleration, µ2MA, the time 293 from maximum acceleration to maximum velocity, µ MA2MV, and the time from onset to MGA, µ2MGA. 294 As the outcome variables are centred, the β values are expressed as a proportion of the measure's 295 overall SD, e.g. µ ma is expressed as a proportion of the overall SD of MA (SDma). .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 19 296 Table 4. MLM estimates for fixed and random effects predicting movement planning of the RGL task. Estimate Max. acceleration Max. velocity MGA Path length Time from onset to max. acceleration Time from max. acceleration to max. velocity Time from onset to maximum grip aperture Intercept −0.77*** [−0.99, −0.54] −1.09*** [−1.24, −0.94] 3.20*** [2.79, 3.61] 0.88*** [0.77, 0.98] 1.89*** [1.48, 2.30] 3.22*** [2.87, 3.56] 3.77*** [3.33, 4.21] Three-way interactions (β) Max. acceleration Max. velocity MGA Path length Time from onset to max. acceleration Time max. acceleration to the max. velocity Time from onset to maximum grip aperture Binocular blur vs. monocular blur * Contrast * Near vs. mid distance - - 0.22* [0.01, 0.43] - NS - NS Monocular blur vs. full vision * Contrast * Near vs. mid distance - - 0.24* [0.03, 0.45] - NS - NS Binocular blur vs. monocular blur * Contrast * Near vs. far distance - - 0.27* [0.06, 0.48] - 0.44* [0.07, 0.81] - 0.39* [0.05, 0.72] Monocular blur vs. full vision * Contrast * Near vs. far distance - - 0.39*** [0.18, 0.60] - NS - NS .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 20 Two-way interactions (β) Max. acceleration Max. velocity MGA Path length Time from onset to max. acceleration Time max. acceleration to the max. velocity Time from onset to maximum grip aperture Binocular blur vs. monocular blur * Contrast NS NS NS NS NS NS NS Monocular blur vs. full vision * Contrast NS NS −0.21** [−0.36, −0.07] NS NS NS NS Binocular blur vs. monocular blur * Near vs. mid distance - 0.06* [0.00, 0.12] −0.16* [−0.31, −0.01] - NS - NS Monocular blur vs. full vision * Near vs. mid distance - NS NS - NS - NS Binocular blur vs. monocular blur * Near vs. far distance - 0.06* [0.00, 0.12] NS - NS - NS Monocular blur vs. full vision * Near vs. far distance - NS NS - −0.41** [−0.67, −0.15] - NS Contrast * Near vs. mid distance - - 0.13*** [0.05, 0.20] - NS - NS Contrast * Near vs. far distance - - NS - NS - NS .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 21 Main effects (β) Max. acceleration Max. velocity MGA Path length Time from onset to max. acceleration Time max. acceleration to the max. velocity Time from onset to maximum grip aperture Binocular blur vs. monocular blur 0.25** [0.08, 0.41] NS 0.55*** [0.31, 0.78] 0.05* [0.01, 0.09] NS NS NS Monocular blur vs. full vision 0.22*** [0.09, 0.35] NS 0.43*** [0.26, 0.60] 0.04* [0.01, 0.07] NS 0.30* [0.06, 0.54] NS Contrast −0.05** [−0.08, −0.01] −0.03** [−0.04, −0.01] −0.10*** [−0.15, −0.05] −0.01* [−0.02, 0.00] NS NS NS Near vs. mid distance 0.88*** [0.74, 1.03] 1.17*** [1.09, 1.25] NS 1.19*** [1.17, 1.20] 0.23* [0.01, 0.45] NS 0.45*** [0.25, 0.65] Near vs. far distance 1.52*** [1.29, 1.75] 2.13*** [1.99, 2.26] NS 2.36*** [2.34, 2.38] NS NS 0.66*** [0.40, 0.92] Stereoacuity NS −0.05* [−0.09, −0.01] 0.11* [0.03, 0.20] NS NS NS NS CS −0.05* [−0.10, 0.00] NS NS 0.03*** [0.02, 0.04] 0.19*** [0.10, 0.27] NS 0.08* [0.01, 0.16] .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 22 VA NS NS 0.25** [0.09, 0.41] 0.04* [0.01, 0.07] NS NS NS Session number NS NS NS −0.01*** [−0.02, −0.01] NS NS NS Random effects (σ) Max. acceleration Max. velocity MGA Path length Time to the max. acceleration Time max. acceleration to the max. velocity Time from onset to the maximum grip aperture Participant ID 0.48 0.32 0.23 0.05 0.4 0.29 0.28 Near vs. mid distance 0.29 0.17 0.09 - 0.22 0.23 0.18 Near vs. far distance 0.49 0.3 0.15 - 0.29 0.34 0.24 Session number 0.1 0.1 0.08 - 0.15 0.11 0.13 Model fit Max. acceleration Max. velocity MGA Path length Time from onset to max. acceleration Time max. acceleration to the max. velocity Time from onset to maximum grip aperture R2 marginal 0.368 0.747 0.49 0.986 0.158 0.236 0.524 R2 conditional 0.852 0.954 0.846 0.989 0.559 0.698 0.885 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 23 Note: All outcome variables and numeric predictors (VA, CS, stereoacuity, and session number) were standardised (i.e., with a mean equal to zero and SD equal to one). Normally distributed variables (maximum velocity and maximum acceleration) were modelled using a Gaussian distribution. In contrast, the remaining variables were modelled using a Gamma distribution. The Gamma only accepts positive values; these variables, therefore, are further transformed to be made positive (MGA + 3, path length + 2, time from onset to the max. acceleration + 2, time from max. acceleration to the max. velocity + 3, and time from onset to the maximum grip aperture + 4). NS = p ≥ .05 * = p < .05, ** = p < .01, *** = p < .001. The numbers in each cell represent the estimate [lower 95% CI, upper 95% CI]. Blank cells (-) indicate that this predictor was not included in the final model. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 24 298 Maximum acceleration 299 MLM analysis revealed significant main effects of visual condition and target contrast on maximum 300 acceleration. Maximum acceleration increased progressively from binocular to monocular blur 301 (67.09mm/s 2 higher) to full vision (59.34 mm/s2 higher) and by 12.54 mm/s2 with high contrast 302 compared to low contrast targets. 303 The SD of the random intercept was 62.34% of the fixed intercept, reflecting moderate variability in 304 baseline MA across participants. 305 Maximum velocity 306 There were two significant two-way interactions between visual condition (binocular blur vs. 307 monocular blur) x distance (near vs. mid) and visual condition (binocular blur vs. monocular blur) x 308 distance (near vs. far). The increase in maximum velocity associated with monocular blur relative to 309 binocular blur was greater at mid and far distances than near distances (see Figure 5). A significant 310 main effect of contrast revealed that maximum velocity increased by 6.85mm/s with high contrast 311 compared to low contrast targets. 312 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 25 313 Figure 5. The effect of distance and visual condition (monocular and binocular blur) on maximum 314 velocity (m/s). Error bars represent standard error (SE). 315 [INSERT FIGURE 5 HERE] 316 Participants' baseline levels of MV showed variability, with the SD of the random intercepts estimated 317 as 29.36% of the fixed intercept. 318 Maximum Grip Aperture (MGA) 319 MLM analysis revealed four significant three-way interactions. When considering the interactions of 320 visual condition (binocular blur vs. monocular blur) x contrast x distance, at both mid and far 321 distances, monocular blur (relative to binocular blur) is associated with a decrease in MGA regardless 322 of target contrast. However, at a near distance, monocular blur (relative to binocular blur) is 323 associated with a reduction in MGA with low contrast targets and a slight increase or no difference in 324 MGA with high contrast targets (Figure 6). Furthermore, when considering the interactions of visual 325 condition (monocular blur vs. full vision) x contrast x distance, at both mid and far distances, full 326 vision (relative to monocular blur) is associated with an increase in MGA regardless of target contrast. 327 However, at near distance, full vision (relative to monocular blur) is associated with increased MGA 328 with high contrast targets and a slight increase or no difference in MGA with low contrast targets 329 (Figure 6). 330 Figure 6. The effect of visual condition (binocular blur vs. monocular blur and monocular blur vs. full 331 vision), contrast and distance (near vs. mid and near vs. far) on MGA (mm). Error bars represent SE . 332 [INSERT FIGURE 6 HERE] 333 MLM analysis revealed significant main effects of visual condition and contrast on MGA. MGA 334 increased progressively from binocular to monocular blur (4 mm larger) to full vision (3.16 mm 335 larger) and by 0.74 mm with high contrast compared to low contrast targets. 336 The SD of the random intercepts was 115.00% of the fixed intercept, indicating high individual 337 variability in base levels of MGA. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 26 338 Path length 339 MLM analysis revealed two significant main effects of visual condition on path length. Path length 340 increased progressively from binocular to monocular blur (4.52 mm longer) to full vision (3.65 mm 341 longer) and by 1.07 mm with high contrast compared to low contrast targets. 342 Time to Maximum Acceleration (OtoMA) 343 A significant three-way visual condition (binocular blur vs. monocular blur) x contrast x distance 344 (near vs. far) interaction indicates that at near distance, with low contrast targets, monocular blur is 345 associated with a slight decrease in OtoMA; however, at far distance with low contrast targets, 346 monocular blur is associated with a slight increase in OtoMA. Monocular blur is not associated with a 347 change in OtoMA with high contrast targets at either distance (see Figure 7). 348 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 27 349 Figure 7. The effect of visual condition (binocular vs. monocular blur), contrast and distance (near vs. 350 far) on the time from onset to maximum acceleration (% of total movement time). Error bars represent 351 SE. 352 [INSERT FIGURE 7 HERE] 353 The baseline level of OtoMA showed substantial variability across participants, with the SD of the 354 random intercepts estimated at 363.64% of the fixed intercept. 355 Time from maximum acceleration to maximum velocity (MAtoMV) 356 A significant main effect of visual condition (monocular blur vs. full vision) indicated that full vision 357 increased MAtoMV by 2.26% compared to monocular blur. 358 The SD of the random intercepts was 131.82% of the fixed intercept, indicating high inter-individual 359 variability in baseline levels of MAtoMV. 360 Time from Onset to MGA (OtoMGA) 361 A significant three-way visual condition (binocular blur vs. monocular blur) x contrast x distance 362 (near vs. far) indicated that monocular blur was associated with a slight increase in OtoMGA at near 363 distance with high contrast targets and at far distance with low contrast targets, whereas, monocular 364 blur was associated with a slight decrease in OtoMGA at near distance with low contrast targets and 365 no apparent change in OtoMGA at far distance with a high contrast target (see Figure 8). 366 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 28 367 Figure 8. The effect of visual condition (binocular blur vs. monocular blur), contrast and distance 368 (near vs. far distance) on the time to maximum grip aperture (% of total movement time). Error bars 369 represent SE. 370 [INSERT FIGURE 8 HERE] 371 Similar to MGA, baseline levels of OtoMGA also showed high individual differences (SD of the 372 random intercepts equal to 121.74% of the fixed intercept). 373 Online control 374 The β values in Table 5 predict maximum deceleration, µ md, the time from maximum velocity to 375 maximum deceleration, µ MV2MD, the time from maximum deceleration to contact with the target, 376 µMD2C, and dwell time, µDwT. As the outcome variables are centred, the β values are expressed as a 377 proportion of the measure's overall SD, e.g., µ md is expressed as a proportion of the overall SD of MD 378 (SD md). .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 29 379 Table 5. MLM estimates for fixed and random effects predicting online control characteristics of the 380 RGL task. Estimate Max. declaration Time from max. velocity to the max. deceleration Time from max. deceleration to contact Time from contact to lift (dwell time) Intercept 0.79*** [0.59, 1.00] 3.29*** [2.90, 3.68] 2.55*** [2.24, 2.86] 2.21*** [1.76, 2.66] Two-way interactions (β) Max. declaration Time max. velocity to the max. deceleration Time from max. deceleration to contact Time from contact to lift Binocular blur vs. monocular blur * Contrast NS NS NS NS Monocular blur vs. full vision * Contrast NS NS NS NS Binocular blur vs. monocular blur * Near vs. mid distance - NS NS - Monocular blur vs. full vision * Near vs. mid distance - −0.33** [−0.57, −0.10] 0.29** [0.08, 0.50] - Binocular blur vs. monocular blur * Near vs. far distance - NS NS - Monocular blur vs. full vision * Near vs. far distance - NS NS - Contrast * Near vs. mid distance 0.11* [0.02, 0.19] NS - - Contrast * Near vs. far distance NS −0.24** [−0.40, −0.08] - - .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 30 Main effects (β) Max. declaration Time max. velocity to the max. deceleration Time from max. deceleration to contact Time from contact to lift Binocular blur vs. monocular blur NS NS NS −0.53*** [−0.78, −0.29] Monocular blur vs. full vision NS NS NS −0.39*** [−0.57, −0.20] Contrast NS NS NS NS Near vs. mid distance −0.99*** [−1.15, −0.83] NS 0.51*** [0.28, 0.74] −0.15 [−0.21, −0.09]*** Near vs. far distance −1.50*** [−1.73, −1.27] −0.38** [−0.67, −0.10] 1.01*** [0.68, 1.33] −0.26*** [−0.32, −0.20] Stereoacuity −0.09* [−0.17, −0.01] −0.15* [−0.29, 0.00] NS NS CS NS NS NS NS VA NS NS NS −0.19** [−0.33, −0.05] Session number NS NS 0.08*** [0.05, 0.12] −0.11*** [−0.14, −0.09] Random effects (σ) Max. deceleration Time max. velocity to the max. deceleration Time max. deceleration to contact Time from contact to lift Participant ID 0.43 0.32 0.25 0.37 Near vs. mid distance 0.32 0.22 0.22 - Near vs. far distance 0.48 0.26 0.31 - Session number 0.12 0.08 - - .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 31 Model fit Max. declaration Time max. velocity to the max. deceleration Time from max. deceleration to contact Time from contact to lift R2 marginal 0.381 0.406 0.559 0.203 R2 conditional 0.817 0.712 0.787 0.614 Note: All outcome variables and numeric predictors (VA, CS, stereoacuity, and session number) were standardised (i.e., with a mean equal to zero and SD equal to one). Maximum deceleration was normally distributed and was modelled using a Gaussian distribution. In contrast, the remaining variables were modelled using a Gamma distribution. The Gamma distribution only accepts positive values; therefore, these variables are further transformed to be positive (time from maximum velocity to maximum deceleration + 3, time from maximum deceleration to contact + 3 and time from contact to lift +2). None of the final models for the variables presented in this table contained a significant three-way interaction, so three-way interactions are not included. The numbers in each cell represent the estimate [lower 95% CI, upper 95% CI]. Blank cells (-) indicate that this predictor was not included in the final model. NS = p ≥ .05 * = p < .05, ** = p < .01, *** = p < .001. 381 Maximum deceleration 382 No significant interactions or main effects were relevant to the study's aims. 383 Time from maximum velocity to maximum deceleration (MVtoMD) 384 A significant two-way visual condition (monocular blur vs. full vision) x distance (near vs. mid) 385 interaction indicates that full vision is associated with an increase in MVtoMD at a near distance and a 386 decrease at a mid distance (see Figure 9). 387 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 32 388 Figure 9. The effect of visual condition (monocular blur vs. full vision) x distance (near vs. mid) on 389 the time from maximum velocity to maximum deceleration (% of total movement time). Error bars 390 represent SE. 391 [INSERT FIGURE 9 HERE] 392 The SD of the random intercepts for MVtoMD showed high individual variability (110.34% of the 393 fixed intercept). 394 395 396 397 398 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 33 399 Time from maximum deceleration to contact with the target (MDtoC) 400 A significant two-way visual condition (monocular blur vs. full vision) x distance (near vs. mid 401 distance) interaction indicated that full vision was associated with a decrease in MDtoC, relative to 402 monocular blur, at a near distance and an increase in MDtoC at a mid distance (see Figure 10). 403 Figure 10. The effect of visual condition (monocular blur vs. full vision) and distance (near vs. mid) 404 on the time from maximum deceleration to contact (% of total movement time). Error bars represent 405 SE. 406 [INSERT FIGURE 10 HERE] 407 The SD of the random intercepts was 55.56% of the fixed intercept, indicating that the base level of 408 MDtoC was moderately heterogeneous between participants. 409 Time from contact to lift (dwell time [DwT]) 410 MLM analysis revealed significant main effects of visual condition on DwT, which decreased 411 progressively from binocular to monocular blur (5.08% lower) to full vision (3.68% lower) and by 412 12.54 mm/s 2 and showed high individual differences in its baseline levels as indicated by the SD of 413 the random intercepts being equal to 176.19% of the fixed intercept. 414 415 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 34 416 Movement time 417 The β values in Table 6 predict movement time, µ mt. As movement time is centred, the β values are 418 expressed as a proportion of movement time's overall SD, SD mt. 419 Table 6. MLM estimates for fixed and random effects predicting online control characteristics of the 420 RGL task. Estimate Movement time Intercept 2.77*** [2.35, 3.19] Three-way interactions (β) Movement time Binocular blur vs. monocular blur * Contrast * Near vs. mid distance NS Monocular blur vs. full vision * Contrast * Near vs. mid distance NS Binocular blur vs. monocular blur * Contrast * Near vs. far distance NS Monocular blur vs. full vision * Contrast * Near vs. far distance -0.23* [-0.45, -0.02] Two-way interactions (β) Movement time Binocular blur vs. monocular blur * Contrast NS Monocular blur vs. full vision * Contrast NS Binocular blur vs. monocular blur * Near vs. mid distance NS Monocular blur vs. full vision * Near vs. mid distance NS Binocular blur vs. monocular blur * Near vs. far distance NS Monocular blur vs. full vision * Near vs. far distance 0.16* [0.01, 0.31] Contrast * Near vs. mid distance NS Contrast * Near vs. far distance NS Main effects (β) Movement time Binocular blur vs. monocular blur −0.27** [−0.47, −0.06] .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 35 Monocular blur vs. full vision −0.34*** [−0.49, −0.19] Contrast NS Near vs. mid distance 0.53*** [0.43, 0.63] Near vs. far distance 1.09*** [0.96, 1.22] Stereoacuity NS CS NS VA NS Session number NS Random effects (σ) Movement time Participant ID 0.21 Near vs. mid distance 0.09 Near vs. far distance 0.11 Session number 0.1 Model fit Movement time R2 marginal 0.736 R2 conditional 0.933 Note: Movement time and numeric predictors (VA, CS, stereoacuity, and session number) were standardised (i.e., with a mean equal to zero and SD equal to one). Movement time was modelled using a Gamma distribution. The Gamma distribution only accepts positive values; therefore, movement time was further transformed to be positive (movement time + 3). The numbers in each cell represent the estimate [lower 95% CI, upper 95% CI]. Blank cells (-) indicate that this predictor was not included in the final model. NS = p ≥ .05, ** = p < .01, *** = p < .001. 421 422 423 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 36 424 A significant three-way visual condition (monocular blur vs. full vision) x contrast x distance (near vs. 425 far) indicated that full vision was associated with the largest decrease in MT at a far distance with low 426 contrast targets (see Figure 11). There were also two main effects of visual condition: monocular blur 427 reduces movement time by 32.80ms compared to binocular blur, and full vision reduces movement 428 time by 41.93ms relative to monocular blur. 429 Figure 11. The effect of visual condition (monocular blur vs. full vision), contrast and distance (near 430 vs. far) on movement time (s). Error bars represent SE . 431 [INSERT FIGURE 11 HERE] 432 The SD of the random intercepts was equal to 91.30% of the fixed intercept, indicating that the base 433 level of MT was highly heterogeneous between participants. 434 Discussion 435 The present study investigated the effects of monocular and binocular cataract-like visual blur and the 436 interaction between visual blur and target contrast on the kinematic and timing characteristics of a 437 reach-to-grasp task, marking a first in the literature. A priori, it was predicted that reducing visual blur 438 would affect the scaling of the maximum grip aperture (MGA) and overall movement time (MT). 439 These effects would be coupled with increased deceleration time and dwell time (from target contact 440 to lifting the target [DwT]). We also predicted that some, but not all, changes to performance 441 associated with visual blur would be magnified by reducing the contrast of the targets. To match the 442 progression of an individual undergoing cataract surgery, the effect of visual condition was estimated 443 as the effect of monocular blur relative to binocular blur (as per first-eye cataract surgery) and the 444 effect of full vision relative to monocular blur (as per second-eye cataract surgery). 445 To assess the effect of the visual condition on participants' vision, contrast sensitivity (CS), 446 stereoacuity, and visual acuity (VA) were tested under each condition. Modelling confirmed that 447 monocular blur (compared to binocular blur) and full vision (compared to monocular blur) were 448 associated with improvements in all three vision tests. This effect was largest for the transition from .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 37 449 binocular blur to monocular blur, relative to the transition from monocular blur to full vision. The 450 effect of visual condition on CS and VA was substantial (R 2 marginal = 0.668 and 0.366, 451 respectively); however, the effect of visual condition on stereoacuity was minimal (R 2 marginal = 452 0.077; [Cohen, 1988]). At the time of writing, there appears to be no multilevel analysis of the visual 453 outcomes of cataract surgery for comparison. However, these outcomes follow a similar pattern to 454 those of previously published literature. For example, bilateral cataract removal was associated with 455 the largest effect for CS (r 2 = 0.46), followed by VA (r2 = 0.41) and the smallest effect for 456 stereoactuity (r 2 = 0.21; 44). 457 In line with research investigating the effects of amblyopia and monocular vision on prehension 458 (1,6,10,11,14), superior vision (in this case, monocular blur relative to binocular blur, and full vision 459 relative to monocular blur) was associated with a decrease in MT and with increased maximum 460 acceleration (MA). No differences in maximum deceleration (MD) were found. This pattern replicates 461 previous findings showing that control groups exhibit significantly greater MA but similar MD 462 compared to amblyopia patients (14). Full vision relative to monocular blur did not affect maximum 463 velocity (MV), while monocular blur relative to binocular blur increased MV at mid and far distances. 464 A trend towards increased MV at a near distance did not reach significance, likely due to the lower 465 speeds of smaller movements, in accordance with Fitts' Law (45,46), where reduced movement 466 magnitude limits the effect. These findings suggest that the improvement from binocular to monocular 467 blur (analogous to first-eye cataract removal; FES) is sufficient to increase the MV in prehension 468 movements, while the transition from monocular blur to full vision (second-eye cataract removal; 469 SES) appears to provide no additional increase in MV. 470 Although the transition from monocular blur to full vision (SES) resulted in subtler improvements in 471 prehension compared to the more dramatic effects of the transition from binocular to monocular blur 472 (FES), gains in kinematic markers, such as reduced DwT and increased MA (which explain the 473 reduction in movement time despite no change in MV), highlight that SES may be associated with 474 critical benefits and enhancements in safety, efficiency, and independence during daily tasks. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 38 475 The grip aperture (the distance between the thumb and index finger) is planned before movement 476 initiation based on the object's perceived size (47). Therefore, changes to grip apertures due to visual 477 blur likely reflect changes in prehensile movements' planning. In this study, monocular blur relative to 478 binocular blur and full vision relative to monocular blur were associated with increases in MGA of 479 4.00mm and 3.16mm, respectively. These results contradict prior research showing that superior 480 vision leads to no change in MGA in control participants compared to amblyopia patients (Grant et 481 al., 2007) or with binocular vision compared with monocular vision (4,5,8). Similarly, the pattern 482 contrasts with other studies reporting increased MGA with poorer vision, suggesting a greater safety 483 margin (6,8,9). The increase in MGA associated with improved vision found in the present study may 484 result from visual blur impairing the participants' ability to accurately judge the target's distance. 485 Servos et al. (1992) proposed that impaired vision (such as monocular vision compared to binocular 486 vision) leads to participants underestimating target distance, which in turn causes the target to be 487 perceived as smaller, resulting in smaller grip apertures. In the present study, the results suggest that 488 improved vision was associated with increased estimates of target distance and, consequently, 489 increased estimates of target size. This led to a larger MGA under monocular blur compared to 490 binocular blur, and under full vision compared to monocular blur. This effect can be further evidenced 491 by comparing MV with binocular and monocular blur. According to Fitts's law, an increase in 492 estimated distance should be associated with an increase in MV (45,46), which can be seen in Figure 493 5. Therefore, it seems reasonable to infer that participants estimated the target distance to be greater 494 with monocular blur vs binocular blur, and scaled the target size accordingly at movement planning. 495 A point of note regarding the effect of visual condition on MGA is that, while the modelling suggests 496 that monocular blur relative to binocular blur is associated with an increase in MGA, this is not 497 represented in Figure 6, except for high contrast targets at a near distance. We considered the 498 hypothesis that this effect might result from the multilevel structure of the analysis creating false 499 positives or erroneous estimates. The effects of this multilevel structure were investigated by 500 conducting a supplementary analysis (not reported in the results section). The simplest model of the 501 effect of visual condition on MGA was calculated; it only included a main effect of visual condition .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 39 502 and a random intercept for each participant. In this case, neither monocular blur relative to binocular 503 blur nor full vision relative to monocular blur was associated with any change in MGA (p > .05). 504 However, when stereoacuity or VA was added to the model, the main effects of both monocular blur 505 relative to binocular blur and full vision relative to monocular blur emerged, both associated with 506 increases in MGA (p < .001, see 507 https://github.com/willsheppard9895/blurNprehension/mgaModels.html for the model summary). 508 The effect of visual condition only emerged when individual differences in stereoacuity and VA were 509 accounted for. This highlights two important considerations: first, that inter-individual differences in 510 visual functions, such as VA and stereoacuity, evidenced by the analysis of random effects, may be 511 masking the effects of visual impairment - in this case, visual blur. This demonstrates the need to 512 consider each individual rather than applying a “one size fits all” methodology in eye care. Further 513 support comes from the heterogeneity of performance on almost all motor outcomes in the RGL task, 514 demonstrated by the analysis of the random intercepts. Second, this also suggests that other aspects of 515 vision contribute to grip scaling beyond clinical measures. One possible explanation is vergence, 516 which is crucial for grip aperture adjustments in prehension tasks (8). Unlike stereoacuity and VA, 517 vergence behaviour is, at least in part, driven by visual blur, meaning that blur manipulation could 518 impair vergence calibration, having a knock-on effect on depth perception (49) and potentially on 519 visuomotor functions, as the observed changes in MGA strongly suggest. 520 Regarding the time course of prehensile movements, the present study showed that during the 521 early/acceleration phase, monocular blur relative to binocular blur had no significant effect on the 522 proportion of movement time spent reaching maximum acceleration (OtoMA) or transitioning from 523 maximum acceleration to maximum velocity (MAtoMV). Similarly, full vision relative to monocular 524 blur was also associated with no overall change in OtoMA. However, we found a significant 2.26% 525 increase in MAtoMV, suggesting that SES might be associated with improved planning of prehensile 526 movements. The increase in the percentage of overall movement time in the early stages of the 527 movement (OtoMA and MAtoMV) suggests that full vision reduces the number of corrections being 528 made in the later stages of the movement. These results contrast with studies that found no effect or an .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 40 529 absolute decrease in acceleration time associated with superior vision (see 14 for an example). On the 530 other hand, the increase in MAtoMV observed here is comparable in scale to previous findings when 531 expressed as a percentage of total movement time, 1.62% (9). 532 In the late/deceleration phase, there were no main effects of visual condition on either the time 533 between max velocity and max deceleration (MVtoMD) or from MD to first contact with the target 534 (MDtoC). The impact of visual condition interacted with near versus mid distance, whereby full 535 vision relative to monocular blur was associated with an increase in MVtoMD at a near distance and a 536 decrease at mid distance (see Figure 9), compared with a decrease in MDtoC at a near distance and an 537 increase in MDtoC at mid distance (see Figure 10). However, these effects are small (~1% difference) 538 and not coupled with a main effect, so they seem inconsequential practically. As reducing visual blur 539 has a limited effect on the time course of the late/deceleration phase of the movement, this suggests 540 that reducing visual blur is associated with lesser changes to the online control of prehensile 541 movements (50) compared to the significant impact of reducing visual blur on metrics associated with 542 movement planning (such as MA, MV and MGA). 543 Although there was no main effect of visual condition on the time from onset to MGA (OtoMGA), a 544 significant three-way interaction suggested that full vision (relative to monocular blur) increased 545 OtoMGA under the most challenging conditions (far distance, low contrast). However, considering 546 the variability in previous findings from the literature (-7.80% to +5.11% change in OtoMGA with 547 binocular vs. monocular vision) and the overall positive effects of full vision on movement planning 548 (e.g., increased MA, MV, and reduced dwell time), this isolated interaction is unlikely to indicate a 549 meaningful impairment in planning. 550 As per previous research, monocular blur and full vision (relative to binocular and monocular blur, 551 respectively) were associated with decreased dwell time (9,10,14). A decrease in dwell time suggests 552 that improving vision improves the representation of the object's position and physical parameters, 553 reducing the reliance on somatosensory feedback from the finger and thumb and allowing the lift to 554 be executed more quickly, analogous to reduced online control errors (10). This result is particularly 555 compelling when coupled with the increases in MA (monocular blur and full vision), MV (full vision .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 41 556 only), and MGA (monocular blur and full vision). It seems reasonable to conclude that reducing 557 visual blur leads to better planning and online control of prehensile movements. Reducing visual blur 558 allows individuals to plan better prehensile movements, most likely due to changes in the perception 559 of the shape/size of the target. This is coupled with an improved ability to make in-flight adjustments, 560 facilitating the efficient and safe performance of the movement. 561 Future studies could benefit from using liquid crystal glasses to control the precise timing of visual 562 input. Grant and Conway (2019) demonstrated that removing visual feedback after movement 563 initiation removed any binocular advantage associated with full vision compared to amblyopia; it 564 would be beneficial to know if this also applies to visual blur. This, in turn, would help us better 565 understand the situations in which patients with conditions such as cataracts may face the greatest 566 difficulties with prehension. Furthermore, it would be good to manipulate visual cues to delineate 567 effects outside of CS, VA and stereoacuity (the effect of visual blur on stereoacuity was very small, R 2 568 = 0.077, see Table 3). For example, Loftus et al. (2004) manipulated the availability of vergence 569 information, allowing us to test the theory suggested earlier in the discussion. Furthermore, future 570 research may want to manipulate blur levels to gain an estimate of the point at which visual blur and 571 visual conditions such as cataracts may begin to make prehensile tasks increasingly difficult. 572 A critical limitation of this study is that our setup did not effectively control for monocular depth 573 cues: without a chinrest, participants might compensate for visual blur by adjusting their head position 574 and viewpoint (although this is not entirely atypical, see 9). This ability to adapt is directly relevant to 575 real-world cataract patients, particularly those who have undergone FES and retain one unimpaired 576 eye. FES patients can partially mitigate the functional limitations of monocular blur through 577 compensatory head movements and behavioural adjustments, reducing their reliance on degraded 578 stereo and vergence cues. However, this compensatory strategy does not replicate the experience of 579 true binocular clarity. The results of this study reinforce that even when individuals have access to 580 monocular adaptation strategies, movement planning and control continue to improve with full 581 binocular vision. SES provides functional advantages that extend beyond behavioural compensation, 582 reinforcing its clinical importance. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 42 583 The results proposed here, along with previously published work, suggest that reducing visual blur 584 and, by proxy, timely FES and SES improve the planning and, to a lesser extent, the online control of 585 prehensile movements. These findings motivate a future programme of work to investigate the effects 586 of FES and SES in cataract patients as they move through their surgery journey, thus providing a 587 better understanding of the potential benefits of cataract surgery on skilled action. 588 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 43 589 Acknowledgements 590 Thanks to everyone that helped. 591 References 592 1. Grant S, Conway ML. Reach-to-precision grasp deficits in amblyopia: Effects of object contrast 593 and low visibility. Vision Res. 2015;114:100–10. 594 2. Crawford JD, Medendorp WP, Marotta JJ. Spatial Transformations for Eye–Hand Coordination. J 595 Neurophysiol. 2004 Jul;92(1):10–9. 596 3. Flanagan JR, Bowman MC, Johansson RS. Control strategies in object manipulation tasks. Curr 597 Opin Neurobiol. 2006 Dec 1;16(6):650–9. 598 4. Bradshaw MF, Elliott KM, Watt SJ, Hibbard PB, Davies IRLL, Simpson PJ. Binocular cues and 599 the control of prehension. Spat Vis. 2004;17(1–2):95–110. 600 5. Gnanaseelan R, Gonzalez DA, Niechwiej-Szwedo E. Binocular advantage for prehension 601 movements performed in visually enriched environments requiring visual search. Front Hum 602 Neurosci. 2014 Nov;8:1–11. 603 6. Grant S, Conway ML. Some binocular advantages for planning reach, but not grasp, components 604 of prehension. Exp Brain Res. 2019;237:1239–55. 605 7. Jackson SR, Jones CA, Newport R, Pritchard C. A kinematic analysis of goal-directed prehension 606 movements executed under binocular, monocular, and memory-guided viewing conditions. Vis 607 Cogn. 1997;4(2):113–42. 608 8. Loftus A, Servos P, Goodale MA, Mendarozqueta N, Mon-Williams M. When two eyes are better 609 than one in prehension: monocular viewing and end-point variance. Exp Brain Res. 2004 Oct 610 1;158(3):317–27. 611 9. Melmoth DR, Grant S. Advantages of binocular vision for the control of reaching and grasping. 612 Exp Brain Res. 2006 May 1;171(3):371–88. 613 10. Buckley JG, Pacey IE, Panesar GK, Scally A, Barrett BT. Prehension of a Flanked Target in 614 Individuals With Amblyopia. Invest Ophthalmol Vis Sci. 2015 Nov 30;56(12):7568–80. 615 11. Grant S, Melmoth DR, Morgan MJ, Finlay AL. Prehension deficits in amblyopia. Invest 616 Ophthalmol Vis Sci. 2007 Mar;48(3):1139–48. 617 12. Grant S, Suttle C, Melmoth DR, Conway ML, Sloper JJ. Age- and Stereovision-Dependent Eye– 618 Hand Coordination Deficits in Children With Amblyopia and Abnormal Binocularity. Invest 619 Ophthalmol Vis Sci. 2014 Sep 8;55(9):5687–701. 620 13. Melmoth DR, Finlay AL, Morgan MJ, Grant S. Grasping Deficits and Adaptations in Adults with 621 Stereo Vision Losses. Invest Ophthalmol Vis Sci. 2009 Aug 1;50(8):3711–20. 622 14. Niechwiej-Szwedo E, Goltz HC, Chandrakumar M, Hirji Z, Crawford JD, Wong AMF. Effects of 623 Anisometropic Amblyopia on Visuomotor Behavior, Part 2: Visually Guided Reaching. Invest 624 Ophthalmol Vis Sci. 2011 Feb 9;52(2):795–803. .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 44 625 15. Niechwiej-Szwedo E, Kennedy SA, Colpa L, Chandrakumar M, Goltz HC, Wong AMF. Effects 626 of Induced Monocular Blur versus Anisometropic Amblyopia on Saccades, Reaching, and Eye- 627 Hand Coordination. Invest Ophthalmol Vis Sci. 2012 Jul 2;53(8):4354–62. 628 16. Suttle CM, Melmoth DR, Finlay AL, Sloper JJ, Grant S. Eye–Hand Coordination Skills in 629 Children with and without Amblyopia. Invest Ophthalmol Vis Sci. 2011 Mar 29;52(3):1851–64. 630 17. Kotecha A, Fahy R, Rubin G. The effects of central and peripheral visual losses on reaching and 631 grasping. Invest Ophthalmol Vis Sci. 2013 Jun 16;54(15):2186. 632 18. Pardhan S, Gonzalez-Alvarez C, Subramanian A. How does the presence and duration of central 633 visual impairment affect reaching and grasping movements? Ophthalmic Physiol Opt. 634 2011;31(3):233–9. 635 19. Pardhan S, Gonzalez-Alvarez C, Subramanian A, Chung STL. How Do Flanking Objects Affect 636 Reaching and Grasping Behavior in Participants with Macular Disorders? Invest Ophthalmol Vis 637 Sci. 2012 Sep 28;53(10):6687–94. 638 20. Pardhan S, Scarfe A, Bourne R, Timmis M. A Comparison of Reach-to-Grasp and Transport-to- 639 Place Performance in Participants With Age-Related Macular Degeneration and Glaucoma. 640 Invest Ophthalmol Vis Sci. 2017 Mar 10;58(3):1560–9. 641 21. Timmis MA, Pardhan S. The Effect of Central Visual Impairment on Manual Prehension When 642 Tasked with Transporting-to-Place an Object Accurately to a New Location. Invest Ophthalmol 643 Vis Sci. 2012 May 14;53(6):2812–22. 644 22. Corveleyn X, Lenoble Q, Rouland JF, Boucart M. What is the Nature of the Reach-and-Grasp 645 Deficit in Glaucoma? J Glaucoma. 2020 Sep;29(9):799. 646 23. Kotecha A, O’Leary N, Melmoth D, Grant S, Crabb DP. The Functional Consequences of 647 Glaucoma for Eye–Hand Coordination. Invest Ophthalmol Vis Sci. 2009 Jan 1;50(1):203–13. 648 24. Whitaker HL, O’Leary N, Crabb DP, Smith ND, Kotecha A. Reaching And Grasping In 649 Glaucoma: The Effect Of Visual Field Defect Location. Invest Ophthalmol Vis Sci. 2011 Apr 650 22;52(14):5504. 651 25. Hess RF, Howell ER. The threshold contrast sensitivity function in strabismic amblyopia: 652 Evidence for a two type classification. Vision Res. 1977 Jan 1;17(9):1049–55. 653 26. Levi DM, Harwerth RS. Contrast sensitivity in amblyopia due to stimulus deprivation. Br J 654 Ophthalmol. 1980 Jan 1;64(1):15–20. 655 27. Levi DM, Harwerth RS. Spatio-temporal interactions in anisometropic and strabismic amblyopia. 656 Invest Ophthalmol Vis Sci. 1977;16(1):90–5. 657 28. Li RW, So K, Wu TH, Craven AP, Tran TT, Gustafson KM, et al. Monocular blur alters the 658 tuning characteristics of stereopsis for spatial frequency and size. R Soc Open Sci. 2016 659 Sep;3(9):160273. 660 29. The Royal College of Ophthalmologists. Changes in NHS cataract surgery in England 2016- 661 2021: an analysis of national, regional and independent sector trends [Internet]. London: The 662 Royal College of Ophthalmologists; 2022 p. 1–19. Available from: 663 https://www.rcophth.ac.uk/wp-content/uploads/2022/08/Cataract-Sector-Data-Report.pdf .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 45 664 30. Bailey IL, Lovie JE. New Design Principles for Visual Acuity Letter Charts. Solutions. 665 1976;53(11):740–5. 666 31. Walraven J. TNO tests stereo acuity (in Dutch). Innovatie. 1973;12(6). 667 32. Walraven J. TNO tests stereo acuity (in Dutch). TNO Proj. 1974;18–9. 668 33. Rombouts SARB, Barkhof F, Sprenger M, Jaap Valk, Scheltens P. The functional basis of ocular 669 dominance: functional MRI (fMRI) findings. Neurosci Lett. 1996 Dec 27;221(1):1–4. 670 34. Maas C, Snijders T. The Multilevel Approach to Repeated Measures for Complete and 671 Incomplete Data. Qual Quant. 2003 Feb 1;37:71–89. 672 35. Campagnoli C. Kinesis: R package for the analysis of motion trajectories and movement 673 kinematics [Internet]. 2024. Available from: https://github.com/ccampagnoli/kinesis 674 36. R Core Team. R: A language and environment for statistical computing [Internet]. Vienna, 675 Austria: R Foundation for Statistical Computing; 2023. Available from: https://www.R- 676 project.org/. 677 37. Bauldry S. Structural Equation Modeling. In: Wright JD, editor. International Encyclopedia of the 678 Social & Behavioral Sciences (Second Edition) [Internet]. Oxford: Elsevier; 2015 [cited 2023 Sep 679 7]. p. 615–20. Available from: 680 https://www.sciencedirect.com/science/article/pii/B9780080970868440559 681 38. Raftery AE. Bayesian Model Selection in Social Research. Sociol Methodol. 1995;25:111–63. 682 39. Bolger N, Zee KS, Rossignac-Milon M, Hassin RR. Causal processes in psychology are 683 heterogeneous. J Exp Psychol Gen. 2019 Apr;148(4):601–18. 684 40. Bates D, Maechler M, Bolker B, Walker S. Fitting Linear Mixed-Effects Models Using lme4. J 685 Stat Softw. 2015;67:1–48. 686 41. Kuznetsova A, Brockhoff PB, Christensen RHB. lmerTest Package: Tests in Linear Mixed 687 Effects Models. J Stat Softw. 2017;82(13):1–26. 688 42. Lenth RV. emmeans: Estimated Marginal Means, aka Least-Squares Means [Internet]. 2017 689 [cited 2024 Nov 29]. p. 1.10.5. Available from: https://CRAN.R-project.org/package=emmeans 690 43. Cohen J. Statistical Power Analysis for the Behavioral Sciences. Second. Hillsdale, NJ: Laurence 691 Erlbaum Associates; 1988. 579 p. 692 44. Gia To K, Meuleners L, Bulsara M, Fraser M, Van Duong D, Van Do D, et al. A longitudinal 693 cohort study of the impact of first- and both-eye cataract surgery on falls and other injuries in 694 Vietnam. Clin Interv Aging. 2014;9:743–51. 695 45. Fitts PM. THE INFORMATION CAPACITY OF THE HUMAN MOTOR SYSTEM IN 696 CONTROLLING THE AMPLITUDE OF MOVEMENT. J Exp Psychol. 1954;47(6):281–391. 697 46. Fitts PM, Peterson JR. Information capacity of discrete motor responses. J Exp Psychol. 1964 698 Feb;67(2):103–12. 699 47. Jeannerod M. The timing of natural prehension movements. J Mot Behav [Internet]. 1984 Sep 700 [cited 2025 Feb 20];16(3). Available from: https://pubmed.ncbi.nlm.nih.gov/15151851/ .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint 46 701 48. Servos P, Goodale MA, Jakobson LS. The role of binocular vision in prehension: a kinematic 702 analysis. Vision Res. 1992;32(8):1513–21. 703 49. Horwood AM. When Does Blur Matter? A Narrative Review and Commentary. J Binocul Vis 704 Ocul Motil. 2022 Apr 3;72(2):57–68. 705 50. Paulignan Y, Jeannerod M, MacKenzie C, Marteniuk R. Selective perturbation of visual input 706 during prehension movements. Exp Brain Res. 1991 Nov 1;87(2):407–20. 707 708 709 710 711 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted June 8, 2025. ; https://doi.org/10.1101/2025.06.05.658186doi: bioRxiv preprint