Sources of reduced visual acuity and spectacle treatment options for individuals with Down syndrome: Review of current literature

Abstract

Individuals with Down syndrome are known to have a greater prevalence of ocular conditions such as strabismus, nystagmus, elevated refractive error, poor accommodative function, elevated higher-order optical aberrations and corneal abnormalities. Related to these conditions, individuals with Down syndrome commonly have reduced best-corrected visual acuity at both far and near viewing distances across their lifespan. This review summarises the various optical sources of visual acuity reduction in this population and describes clinical trials that have evaluated alternative spectacle prescribing strategies to minimise these optical deficits. Although refractive corrections may still have limitations in their ability to normalise visual acuity for individuals with Down syndrome, the current literature provides evidence for eye care practitioners to consider in their prescribing practices for this population to maximise visual acuity. These considerations include accounting for the presence of elevated higher-order aberrations when determining refractive corrections and considering bifocal lens prescriptions, even for young children with Down syndrome.

Key points

• This review provides a summary of findings from studies identifying both optical and sensory limitations in best-corrected visual acuity for individuals with Down syndrome.

• Normalising visual acuity for the majority of patients with Down syndrome remains a challenge, but spectacle prescribing strategies that take into account optical deficits offer promise.

• There is evidence that individuals with Down syndrome may benefit from bifocal lenses in childhood to optimise near visual acuity.

PURPOSE

Down syndrome is a congenital condition most commonly from non-disjunction of chromosome 21 prior to or at conception which results in an extra full or partial chromosome (Trisomy 21). Down syndrome causes intellectual disability and developmental delays and can also cause other medical abnormalities, including heart and gastrointestinal disorders. The national prevalence of Down syndrome has continually grown with medical advances and it is currently estimated that 1 in every 528 births in the United States is a child with Down syndrome.1

Individuals with Down syndrome also have a myriad of ocular and visual problems which include high levels of refractive error,2,3 reduced visual acuity,4,5 poor ocular accommodation even in childhood,6,7 thinner and steeper corneas,8,9 increased higher-order wavefront aberrations,10,11 greater prevalence of strabismus12,13 and a greater prevalence of nystagmus.13,14 Each of these conditions alone could have a negative impact on the quality of vision, with an even greater negative impact for individuals who have multiple combined findings. In addition to ocular and visual complications, the intellectual disability and developmental delays commonly found in these individuals may reduce their ability to comprehend the instructions of standard clinical techniques for diagnosis and treatment. Thus, practitioners may be reluctant to examine individuals with Down syndrome and feel that the care they are able to provide is sub-optimal to that available for typically developed individuals without Down syndrome.

Providing the highest level of care possible at an early age is important in maximising visual potential in individuals with Down syndrome. Visual deficits may compound the deficits experienced due to intellectual disability and thus eliminating as many visual barriers as possible is critical to maximising potential in this population. With advances in healthcare, individuals with Down syndrome are living longer lives and thus will likely outlive their parents or guardians on whom they depend. The eye care provider can play an important role in promoting greater independence for these individuals.

While the ocular and visual deficits experienced by this group span a wide variety of conditions, the emphasis of this review will be on deficits that are related to the optical quality of the eye and the desire to maximise visual acuity in this population. In particular, studies investigating refractive error, accommodation and corneal contributions to refractive error and image quality will be reviewed. This review will not address the impact of strabismus or nystagmus on visual acuity in patients with Down syndrome. Although these conditions are commonly found in patients with Down syndrome,12,15 their visual impact extends beyond optical deficits and thus often requires further intervention, such as surgery, for their management. Instead, this review will emphasise the optical deficits prevalent in Down syndrome that may be managed with traditional spectacle lenses. An overview of visual acuity deficits will be provided first, followed by optical findings potentially accounting for these deficits in visual resolution, and lastly, recent work to develop clinical strategies to address visual acuity deficits in this population.

RECENT FINDINGS

Visual acuity

Reduced visual acuity

Reduced visual acuity is a common finding in individuals with Down syndrome, even in the absence of ocular pathology and when corrected with spectacles.5,7 Courage et al. demonstrated that the reduction in visual acuity for individuals with Down syndrome is observed as early as 6 months of age.5 In their study, visual acuity was measured with Teller acuity cards16 on 51 children with Down syndrome aged 2 months to 18 years (median age = 50 months) and findings were compared to developmental norms for individuals without neurological, visual or developmental disorders17 as a function of age. Comparisons showed that 94% of the children with Down syndrome had visual acuity below the age-matched mean and 79% fell more than 2 standard deviations (SD) below the mean, all of whom were older than 6 months of age.5 A sub-analysis was performed on 27 individuals who did not have ocular conditions (e.g., strabismus and nystagmus) and were wearing a refractive correction, but the proportion of individuals with visual acuity falling below 2 SD of the norm remained the same, suggesting that visual acuity does not develop to age normal levels even in the absence of ocular conditions and when corrected with spectacles.5

Similar reports of poor visual acuity in children with Down syndrome have been published by Woodhouse et al.7 In their study, 53 children with Down syndrome aged 12 weeks to 57 months (mean age = 24 months) were tested binocularly with either Teller acuity cards16 (age 12 months and under) or the Cardiff acuity test18 (age greater than 12 months). Visual acuity measures were compared to published age norms from typical children without Down syndrome. The findings of this study demonstrated that visual acuity fell within the normative range for participants less than 2 years of age, but that beyond this age, visual acuity was reduced and showed no continued improvement.7 For participants greater than 2 years of age, there was no difference in acuity between children with and without corrected refractive error, similar to the findings of Courage et al.7

Behavioural limitations and visual acuity

Along with a reduction in visual acuity come questions regarding the source of the acuity deficit. In particular, one investigative group sought to determine whether behavioural limitations from intellectual disability were contributing to the poor acuity observed in children with Down syndrome.19 In their investigation, individuals with Down syndrome from age 9 months to 12 years and controls without Down syndrome from 3 months to 14 years of age were tested both behaviourally with an age-appropriate acuity test (Teller acuity cards, Cardiff acuity test, Kay picture test, logMAR crowded test or the Bailey–Lovie letter chart) and objectively with a steady-state visually evoked potential (VEP) procedure utilising sine-wave grating stimuli. VEP measures were completed successfully on 36 children with Down syndrome and 40 controls.19 Behavioural acuity was obtained for 54 children with Down syndrome and 35 controls. An analysis of covariance with subject group (Down syndrome, control) as the independent variable, acuity as the dependent variable and age as the covariate found that the group with Down syndrome had significantly lower visual acuity than the control group for both the behavioural and VEP acuity measures (p < 0.01).19 These differences were still observed when excluding individuals with Down syndrome with ocular anomalies, such as poor accommodation and nystagmus. A direct comparison of VEP and behavioural acuity for individuals with Down syndrome who completed both tests revealed a statistically significant difference (t₍₃₅₎ = 8.68, p < 0.01) with approximate mean behavioural acuity of 17 ± 7 cycles/degree and VEP acuity of 8 ± 3 cycles/degree.19 Overall, these findings are supportive of a sensory deficit-limiting visual acuity in individuals with Down syndrome rather than a behavioural deficit-limiting testability.

Repeatability of visual acuity testing

Further evidence that individuals with Down syndrome are capable of performing behavioural measures of visual acuity is provided by two studies of adults with Down syndrome to determine the repeatability of Bailey–Lovie style20 visual acuity measures. Ravikumar et al.4 and Anderson et al.11 reported both intra-session repeatability and inter-session repeatability of monocular visual acuity measures in 30 adults with Down syndrome aged 18–50 years4 and 18–52 years11 using a custom, computerised visual acuity system presenting either a 10 letter set (D,E,F,H,N,P,R,U,V,Z) or a 4 letter set (H,O,T,V) arranged in a logMAR style chart with 5 letters per line.

All participants were tested with their presenting correction if they had one and began reading or matching letters with the largest letter size (0.80 logMAR), continuing line by line until a total of five mistakes were made. Letter-by-letter scoring was used to determine the visual acuity measure. For the intra-session repeatability, acuity measures were repeated three times and a group of age-matched controls were included for comparison.4 There was no relationship between the standard deviation and either visual acuity level or participant age.4 Repeatability was 0.13 logMAR (6.5 letters) for participants with Down syndrome, 0.09 logMAR (4.5 letters) for controls for the full letter set and 0.16 logMAR (8 letters) in both groups with the four-letter chart.4 Overall, repeatability was similar for individuals with Down syndrome compared to controls despite the large differences in visual acuity achieved.

In the inter-session repeatability study of adults with Down syndrome, repeatability of binocular visual acuity measures was assessed with visits spaced 19–71 days apart (mean = 35 days).11 For this study, 29 participants were tested with the full 10-letter set and 1 participant was tested with the 4-letter set. The mean difference in binocular visual acuity between visits was 0.02 + 0.06 logMAR (range = −0.10 to 0.14), giving a coefficient of repeatability of 0.12 logMAR (6 letters).11 The difference in acuity was not linearly related to the level of acuity as shown in Figure 1.11 The findings of both of these studies demonstrate that individuals with Down syndrome who are able to complete an initial logMAR style visual acuity test exhibit good repeatability despite their overall reduced visual acuity.

FIGURE 1.

Inter-session comparison of binocular distance visual acuity measures in adults with Down syndrome. The solid line indicates the mean difference and the dashed lines represent the 95% limits of agreement. This figure was previously published by Anderson et al. in 2021.11

Interferometric testing of visual acuity

Given that visual acuity is reduced in individuals with Down syndrome who do not have behavioural limitations to completing visual acuity testing, additional investigations were conducted to identify other significant factors leading to reduced visual acuity. Little et al.21 compared visual acuity measures obtained by grating resolution testing versus interferometric grating testing in both children with Down syndrome and age-matched controls. The premise of this study was that the interferometric testing by-passes the optics of the eye, which may be poor in individuals with Down syndrome, due to their increased risk for abnormalities such as early-onset cataract,22 keratoconus,3 refractive error2,23 and poor accommodative function.7 Thus, improvement in acuity measures with interferometric testing would support an optical limitation to visual acuity. The study sample ranged from 9  to 16 years of age and 26 children with Down syndrome completed interferometric testing and 27 completed the grating resolution acuity test. The control group included 67 children tested with interferometry and 65 children completing the grating resolution test. The grating resolution test results were significantly poorer for children with Down syndrome (mean = +0.48 logMAR ±0.09) than the controls (mean = −0.12 logMAR ±0.07) by two-sample t-test (p < 0.0001).21 However, this difference became markedly smaller when comparing interferometric acuity (Down syndrome mean = +0.003 logMAR ±0.06; control mean = −0.11 logMAR ±0.08).21 Although the visual acuity levels of the individuals with Down syndrome still did not reach that of the controls with interferometric testing, the dramatic improvement suggests that optical deficits are at least in part contributing to reduced visual acuity in this population.

Vernier tests of visual acuity

The same investigative team also measured Vernier acuity thresholds in children with and without Down syndrome as a means of evaluating cortical contributions to visual acuity performance. In comparing grating and Vernier acuity tasks, it is believed that grating detection corresponds to and is limited by retinal ganglion cell density, whereas Vernier acuity depends more on information processed at the cortical level.24,25 In their study, Little et al. obtained monocular Vernier acuity measures on 25 participants with Down syndrome and 65 controls ranging in age from 9 years to 16 years. Vernier acuity was significantly reduced in participants with Down syndrome (mean = 39.8 s of arc) as compared to controls (mean = 14.6 s of arc) by one-way ANOVA (p < 0.001).26 The reduction in acuity for participants with Down syndrome was observed even when excluding the three participants who had nystagmus. The conclusions of this investigative team were that both optical and cortical deficits contributed to visual acuity reduction in individuals with Down syndrome.

Accommodation

Prevalence of reduced accommodation in children with Down syndrome

In addition to experiencing reduced distance visual acuity, individuals with Down syndrome are more likely to suffer poor accommodative function, even in childhood, and thus may experience additive negative effects on visual acuity at near viewing distances. Early reports of poor accommodative function in this population were based on comparisons of distance and near visual acuity, revealing that near visual acuity was dramatically reduced compared to distance measurements, and subsequently improved for some individuals with Down syndrome upon the introduction of plus lenses for near.27 Later studies utilised objective methods to measure accommodative function, such as dynamic retinoscopy at various near viewing distances.7,28 Woodhouse et al.28 reported Nott retinoscopy findings for 24 children with Down syndrome from 6 to 14 years of age, as compared with 26 control children for target distances corresponding to 6, 8 and 11 dioptres (D). The largest accommodative effort exerted in testing these three viewing distances was recorded as the accommodative amplitude. This study found that the majority of controls had values of 10 D or higher, whereas only two participants with Down syndrome showed this level. Thus, 92% of the children with Down syndrome had amplitudes lower than 10 D, with a median of 4.62 D for the group.28 A second study by Woodhouse et al.7 used the same methods to report accommodative findings for even younger children with Down syndrome aged 12 weeks to 57 months and made similar conclusions in that only 4 of the 49 children tested (8.2%) had accommodative responses within the normal range for their age. In addition, there was no significant improvement in accommodative response as a function of age for this cohort.7

Motor versus sensory deficits in accommodation

These findings of a reduced accommodative response in such a large percentage of children with Down syndrome raised the question of whether reduced accommodation was related to deficits in the motor aspects of the accommodative mechanism itself, or to a sensory deficit leading to a reduction in the signal to accommodate. In addition, some of the previous work with Nott retinoscopy included participants who were not wearing a refractive correction, and thus questions remained about whether accommodative performance might be different if individuals were habitually wearing their refractive correction. Anderson et al.6 sought to address these questions with objective measurements of multiple aspects of accommodation. For the study, static measurements of accommodation (accommodative amplitude and accommodative lag) were obtained with the open-field Grand Seiko Autorefractor (formerly manufactured by RyuSyo Industrial Co, Kagawa, Japan), while dynamic measurements of accommodation (latencies, microfluctuations and peak velocities) were obtained with a custom photo-refraction system capable of recording 60 measurements of refraction per second. Participants ranged from 3 to 40 years of age and received a complete eye examination prior to measures of accommodation. Participants whose refractive error was not well corrected were prescribed new spectacles and returned 1 month later for accommodative measures. A cohort of controls without Down syndrome, both children and adults, was also tested for comparison.

Grand Seiko autorefraction was performed first with the target fixed at 33 cm and for subsequent increasing accommodative demands by introducing minus lenses in front of the viewing eye in −1 D steps until no further increase in accommodation could be detected. Individuals with Down syndrome were also tested with proximally induced demands by moving the target physically closer to the positions corresponding to 3, 4, 5, 6, 7 and 8 D. The accommodative amplitude was identified as the largest accommodative response exerted for the series of demands tested (both minus lens induced and proximal) after adjusting responses for the effectivity of any lenses in front of the viewing eye (spectacle and/or trial lenses). Accommodative lag was calculated as the difference between the demand and response for the 33 cm target position and the first five trial lenses (−1 to −5 D).

Maximum accommodative amplitude

Maximum accommodative amplitudes were successfully obtained on 19 individuals with Down syndrome, 9–39 years of age, with an average amplitude of 2.52 ± 1.66 D for the group. Amplitude showed no significant improvement with age or distance visual acuity (p ≥ 0.19) and was overall dramatically lower for individuals with Down syndrome than a cohort of 140 controls without Down syndrome (Figure 2).

FIGURE 2.

Maximum accommodative amplitudes determined from minus lens stimulated demands and open-field Grand Seiko autorefraction as a function of age. The solid line represents a curvilinear fit to the control participant data after excluding the circled points representing young children without Down syndrome who had previously been found to not respond well to minus lens-induced blur. The dashed lines represent ±2 standard deviations surrounding the curvilinear fit of the control participants. This figure was previously published by Anderson et al. in 2011.6

Previous observations in control participants found that some young children do not respond well to minus lens-stimulated accommodative demands (circled points in Figure 2) and thus the individuals with Down syndrome were also evaluated with proximally stimulated accommodative targets; however, the group mean improvement was less than 1 D with an average of 3.30 ± 1.54 D.

Accommodative lag

Figure 3 shows the accommodative lag for five stimulus demands for the same participants who completed the maximum accommodative amplitude testing. There was no age-related change in accommodative lag for the participants with Down syndrome, and thus all participants with Down syndrome were binned together. Accommodative lag was significantly greater at each effective demand for participants with Down syndrome than all of the controls, irrespective of age (one-way ANOVA, p < 0.05).6 The finding that both maximum accommodative amplitude and accommodative accuracy were significantly reduced in a large percentage of individuals with Down syndrome agrees with previous studies,7,28 even with the participants being well corrected and adapted to their refractive corrections prior to accommodation measures.

FIGURE 3.

Accommodative lag measured for a stimulus positioned at 33 cm and five subsequent increasing demands obtained by introducing minus lenses from −1 through −5 D (effective demands calculated to account for introducing lenses in the spectacle plane over the measured eye) as a function of age in years. This figure was previously published by Anderson et al. in 2011.6

Dynamic measures of accommodation

To investigate the question of whether sensory or motor deficits are contributing to reduced accommodation in individuals with Down syndrome, dynamic recordings of accommodation were obtained while participants viewed a cartoon movie alternating between a computer monitor placed at 6 m versus near positions of 2, 3, 4 and 5 D accommodative demands. A minimum of 3 distance/near cycles were presented at each demand in pseudo-random durations to avoid anticipation. Latencies, peak velocities and microfluctuations were obtained from offline analysis of the recordings that had responses suitable for analysis. Of note, only 13.5% of all responses recorded from individuals with Down syndrome had a typical response shape matching the position of the stimulus (Figure 4), whereas 90.8% of responses recorded from controls had a typical response shape. The majority of recordings from individuals with Down syndrome resulted in no response (54.9%) despite the participant's engagement in watching the movie with good fixation. Data from 15 individuals with Down syndrome were able to be analysed.

FIGURE 4.

Accommodative response profiles observed for participants with Down syndrome during dynamic recordings of accommodation to a stimulus alternating between distance and near positions. The step stimulus is shown to represent the accommodative demand of the stimulus over time. The line trace indicates the accommodative response of a control participant (typical response (a)) versus response types of participants with Down syndrome (early terminated (b), atypical (c) and no response (d)). This figure was previously published by Anderson et al. in 2011.6

Overall, approximately 83% of recorded accommodative and disaccommodative latencies (time from stimulus change to the initiation of a change in accommodation) were similar between individuals with and without Down syndrome. The speed of change in accommodation expressed as peak velocities (dioptres/second) was also similar between individuals with and without Down syndrome for both increasing and decreasing accommodation (Figure 5). Microfluctuations, however, were significantly elevated in individuals with Down syndrome (p < 0.001) when compared to controls, as shown in Figure 6.

FIGURE 5.

Accommodative (a) and disaccommodative (b) peak velocities (Vmax) measured in dioptres per second (D/s), plotted as a function of the magnitude (amplitude) of the response. Dashed lines show the mean and 95% limits of predictive fits for control data based on age-matched data and the solid line shows a linear regression fit to the data of participants with Down syndrome. These figures were previously published by Anderson et al. in 2011.6

FIGURE 6.

Accommodative microfluctuations (calculated from the root mean square (RMS)) as a function of accommodative response. Control participants are shown with open symbols and participants with Down syndrome (DS) by filled symbols. Linear regressions were fitted to each data set as shown by the lines and equations. This figure was previously published by Anderson et al. in 2011.6

In total, these findings point towards a sensory deficit leading to reduced accommodative responses. This conclusion is based on beliefs that the peak velocities are driven purely by motor aspects of accommodation, whereas the microfluctuations of accommodation have contributions from both motor and sensory aspects of accommodation.6

Ciliary muscle thickness

Further investigation into potential motor deficits was performed by imaging the ciliary muscle with optical coherence tomography in adults with Down syndrome.29 Ciliary muscle thickness was measured 1, 2 and 3 mm posterior to the scleral spur and compared with previously published datasets of neuro-typical children and adults with both myopic and hyperopic refractive errors (Table 1). Ciliary muscle thickness in adults with Down syndrome was found to be in a similar range to previous study cohorts and demonstrated the same relationships observed previously with refractive error, namely thicker posterior muscles with increasing myopic refractive error and thicker apical fibres with increasing hyperopic refractive error. Thus, there is no evidence that the ciliary muscle differs in individuals with Down syndrome as compared to those without Down syndrome. However, a limitation of this study was that images were obtained post-dilation, and thus it is not possible to say how the anatomy of the ciliary muscle appears during near viewing.

TABLE 1.

Mean (SD) ciliary muscle thickness (CMT) at locations 1 (CMT1), 2 (CMT2) and 3 (CMT3) mm posterior to the scleral spur.

	n	Age (years)	Refractive error spherical equivalent (D)	CMT1 (μm)	CMT2 (μm)	CMT3 (μm)
Adults with Down syndrome29	26	29 (9)	−0.90 (5.03)	804 (83)	543 (131)	312 (100)
			(−15.75 to +5.13)
Adults from Kuchem et al.30	29	28 (6)	−2.56 (3.29)	827 (77)	599 (101)	349 (85)
			(−8.40 to +5.84)
Children from Pucker et al.31	269	8.7 (1.5)	+0.41 (1.29)	809 (68)	528 (73)	281 (55)
			(−4.01 to +7.76)

The finding that ciliary muscle parameters did not differ significantly between individuals with and without Down syndrome was subsequently supported by a study of 16 pre-presbyopic adults with Down syndrome, compared with a sample of 16 age-matched individuals without Down syndrome in whom non-cycloplegic imaging was performed.32

Refractive error

Prevalence of refractive error

There is no doubt that significant refractive error, if left uncorrected during the developmental years, may contribute to the reduced visual acuity observed in individuals with Down syndrome. Many reports have established the high prevalence of refractive error through cross-sectional samples of children and adults with Down syndrome.13,33,34 Refractive error often includes high levels of astigmatism and hyperopia, although high myopia can also be observed. For example, a study of 152 children with Down syndrome aged 2 months to 18 years reported 60% with astigmatism, 28% with hyperopia and 13% with myopia, as determined by cycloplegic retinoscopy.13 This same study reported an association between myopic refractive error and the presence of cardiac malformations. Bromham et al.34 also noted the association between myopic refractive error and cardiac defects. In their sample of 58 children with Down syndrome, 31 had a history of congenital heart defects and all 5 children with myopia had congenital heart defects.34 However, hyperopic refractive error has been demonstrated to be more common in this population overall.3,23,35 Doyle et al. reported non-cycloplegic refractive findings for 50 teenagers and young adults with Down syndrome (age 15–22 years) and observed that 80% of the group had hyperopia in the range from +0.50 to +7.50 D.3 In this sample, only 2% had reached emmetropia (refractive error within ±0.50 D), thus raising the question of whether individuals with Down syndrome undergo emmetropisation in childhood.

Emmetropisation

The hypothesis that individuals with Down syndrome may not undergo emmetropisation led to a desire to study longitudinal changes in refraction. Haugen et al.23 performed >2 years of longitudinal follow-up (average 55 ± 23 months) on 60 infants and toddlers with Down syndrome (age at first examination ranged from 3 to 61 months) and observed a lack of emmetropisation as was suspected from prior cross-sectional studies. In their sample, three different classifications of longitudinal change were observed: Stable hyperopia (n = 34), increasing hyperopia (n = 11) and decreasing hyperopia or the development of a myopic shift (n = 9). Haugen et al. also assessed accommodative function with dynamic retinoscopy in this cohort by examining the retinoscopy reflex from 50 cm with fixation at the retinoscope and then introducing a near target 20–30 cm in front of the child while continuing to perform retinoscopy at 50 cm. A normal accommodative response was classified as a change in the reflex from ‘with’ to ‘against’ when introducing the near target. If this shift was not observed, then the accommodative response was classified as ‘weak’. Interestingly, Haugen et al. observed a relationship between the magnitude of hyperopia and the accommodative response. As shown in Table 2, individuals who had stable, low-grade hyperopia were more likely to have normal accommodative responses. Accordingly, Haugen et al.23 proposed that normal accommodative function may play a role in emmetropisation. This same hypothesis that poor accommodation may lead to higher levels of hyperopia in individuals with Down syndrome has also been proposed previously by Woodhouse et al.35

TABLE 2.

Distribution of normal and weak accommodative responses by refractive error type.

Refractive group	Accommodation
	Normal	Weak	Total
Isometropia
Stable, low-grade hyperopia (≤+2.00 D)	14 (78%)	4 (22%)	18
Stable, moderate-grade hyperopia (+2.25 to +4.00 D)	6 (50%)	6 (50%)	12
Stable, high-grade hyperopia (>+4.00 D)	0	4	4
Increasing hyperopia (≥+1.50 D increase)	4 (36%)	7 (64%)	11
Decreasing hyperopia (≥1.50 D decrease)	2 (22%)	7 (78%)	9
Anisometropia	1 (17%)	5 (83%)	6
Total	27 (45%)	33 (55%)	60 (100%)

Note: A normal accommodative response was defined as a shift in the retinoscopic reflex from ‘with’ to ‘against’ as an examiner performed retinoscopy from 50 cm and the participant switched fixation from 50 cm to a more proximal position of 20–30 cm. A weak accommodative response was the absence of a shift in reflex from ‘with’ to ‘against’ when fixating on the more proximal target. Data from Haugen et al.23

Cregg et al.15 explored the question of emmetropisation with a longitudinal study of 55 children who were first examined at less than 2 years of age and again on at least two more occasions using Mohindra retinoscopy.36 Their findings agreed with those of Haugen et al.23 in that a reduction in hyperopic refractive error was rarely observed. Of the 24 children with a significant refractive error at the initial visit, only six demonstrated emmetropisation while the remainder maintained or increased their refractive error.15 Ten children who were initially emmetropic developed a significant refractive error over the course of the investigation and 21 maintained emmetropia throughout the study.15 However, it should be noted that the definition of emmetropia in this study was more generous than other investigations and included refractive errors with the most ametropic meridian falling between −1.00 and +3.00 D.

Additional evidence for lack of emmetropisation in Down syndrome was provided by Al-Bagdady et al.2 In this retrospective study of 182 individuals with Down syndrome, power vector analysis37,38 was used to compare refractions documented in the medical records of individuals from 6 months through 15 years of age. Individuals were distributed into 1 ± 0.5-year age bins for comparisons, with between 20 and 78 subjects in each bin. Many individuals had been patients in the clinic for multiple years, and thus they may have been included in multiple age bins. Additionally, an analysis of two refractive time points was conducted for individuals seen both at a young age (closest visit to 4 years of age) and during their teenage years (closest visit to 15 years of age). Al-Bagdady et al. reported no significant difference in the mean M component (spherical equivalent) across all 15 age bins, which tended to approximate between +2.00 and +3.00 D (visual inspection of published figure). Astigmatism did show a significant change with age, with the magnitude increasing with age and the axis shifting towards the oblique (shift from 90° to 45°). These trends in refractive error were also observed in the analysis of the 12 individuals with longitudinal follow-up between 4 and 15 years.2

Contributions of corneal astigmatism to refractive astigmatism

Studies have also been conducted to understand the contributions of the optical components of the eye to the refractive error of individuals with Down syndrome. Little et al.39 compared the relationship between refractive astigmatism obtained using Mohindra retinoscopy and corneal astigmatism obtained with hand-held keratometry in 29 children with Down syndrome from 9 to 16 years of age. Analysis was performed after conversion of refractive values to vector notation,37 and the linear relationship was evaluated between corneal and refractive astigmatism for both the horizontal and vertically oriented astigmatic component (J₀) and the obliquely oriented astigmatism component (J₄₅). Surprisingly, there was no significant relationship observed between corneal and refractive astigmatism in participants with Down syndrome, and thus the study concluded that corneal astigmatism could not be used to predict refractive astigmatism in patients with Down syndrome.39

Knowlton et al.9 conducted a similar study in a larger sample by comparing corneal astigmatism obtained from anterior corneal topography with refractive astigmatism measured with an open-field autorefractor to understand the contribution of corneal astigmatism to refractive astigmatism. Their analysis included 128 individuals with Down syndrome aged 8–55 years and 137 controls without Down syndrome aged 7–59 years. In comparing the autorefraction findings, the group mean spherical equivalent refractive error was significantly less myopic in participants with Down syndrome (−0.43 ± 4.03 D) as compared with the controls (−1.31 ± 2.42 D; p = 0.03) and refractive astigmatism was significantly greater in participants with Down syndrome than the controls (−1.94 ± 1.30 DC vs. −0.66 ± 0.60 DC; p < 0.001), as shown in Figure 7a,c. The corneal power measured using topography was also significantly higher in participants with Down syndrome (45.82 ± 1.80 D vs. 43.37 ± 1.58 D; p < 0.001)—See Figure 7b. Both corneal astigmatism and refractive astigmatism were converted using vector analysis37 and internal astigmatism (Figure 7c) was calculated using the equation:

FIGURE 7.

Boxplots for individuals with Down syndrome (DS) compared to controls without Down syndrome showing the distribution of spherical equivalent refractive error as measured by autorefraction (a), corneal power as measured by topography (b) and cylindrical refractive power obtained from autorefraction (expressed as absolute values), corneal toricity obtained by topography and calculated internal astigmatism obtained by subtracting corneal toricity from refractive astigmatism (c). Solid lines represent median values, dashed lines represent mean values, box boundaries represent 25th and 75th percentiles and whiskers represent 10th and 90th percentiles. Figures previously published by Knowlton et al. in 2015.9

There was no significant relationship between the corneal power and spherical equivalent refractive error for the group with Down syndrome (p = 0.91), despite the large range in corneal powers and refractive errors being observed in these participants (Figure 8). There was, however, a strong relationship (r² ranged from 0.31 to 0.74) between corneal and refractive astigmatism for both horizontal and vertically oriented astigmatism (J₀) and oblique astigmatism (J₄₅) as shown in Figure 9. These findings agree with the expectation that corneal astigmatism is the primary contributor to refractive astigmatism, as is seen in the typical population,40 but disagrees with the previous findings of Little et al.39 However, several methodological differences existed between the studies (Mohindra retinoscopy vs. open-field autorefraction for obtaining refractions, and handheld keratometry vs. anterior corneal topography for measuring corneal astigmatism) and the sample in Little et al.39 included only children with a smaller distribution of refractive astigmatism as compared with that of Knowlton et al.9

FIGURE 8.

Linear regression of mean corneal power (measured by topography) versus spherical equivalent refractive power (M) measured by autorefraction. The solid line indicates the regression for the group with Down syndrome (DS) and the dashed line represents the regression for the control group. Figure previously published by Knowlton et al. in 2015.9

FIGURE 9.

Corneal and refractive power vectors were significantly correlated for the J₀ (a) and J₄₅ (b) vectors in participants both with and without Down syndrome. Solid lines indicate correlations for the group with Down syndrome and dashed lines indicate the correlations for the control group. Figures previously published by Knowlton et al. in 2015.9

Cornea

In addition to being a source of refractive astigmatism, the cornea of individuals with Down syndrome may contribute to the reduced optical quality of the eye due to structural differences in this population. For at least 70 years, the literature has included reports of an association between keratoconus and Down syndrome,41 most often based on reports of a small number of cases with advanced corneal disease.42,43 Clinical adoption of corneal topography brought additional studies demonstrating morphologic differences in the cornea of individuals with Down syndrome as compared to those without the condition. Doyle et al.3 reported overt keratoconus in 2% of teenagers included in a cohort of 50 individuals aged 15–22 years and found an additional 6% with corneal topography demonstrating inferior steepening without additional clinical signs of keratoconus.

Corneal power and central corneal thickness

A more quantitative study by Haugen et al.8 performed corneal topography and central corneal thickness measurements on 41 individuals with Down syndrome between 14 and 26 years of age. This study reported detection of keratoconus from topography measures in four of the participants. Mean keratometry power was significantly higher in those with Down syndrome than the 51 controls (46.39 ± 1.95 D vs. 43.41 ± 1.40 D, p < 0.001)8; a finding that agrees well with Knowlton et al.9 as shown in Figure 7b. Central corneal thickness was significantly thinner in individuals with Down syndrome than the controls (480 ± 40 μm vs. 550 ± 30 μm, p < 0.001).8 This finding was also reported by Aslan et al. in a comparison of 27 children with Down syndrome aged 5 to 12 years (corneal thickness = 494 ± 47 μm), with 37 controls without Down syndrome aged 6 to 12 years (corneal thickness = 539 ± 38 μm).44

Keratoconus detection metrics

Vincent et al. used corneal topography to compare 21 children with Down syndrome aged 10 months to 18 years to a group of 60 controls aged 5–15 years and also found elevated corneal power in the group with Down syndrome (46.66 ± 1.64 D vs. 42.60 ± 1.87, p < 0.001).45 In addition, the investigators calculated Inferior–Superior corneal power values as a means of diagnosing eyes suspected of having keratoconus. Unfortunately, Inferior–Superior calculations could only be performed on 12 eyes due to poor-quality images from the upper eyelid blocking the mires. Of the 12 eyes evaluated, two had an Inferior–Superior value with significant inferior steepening. Interestingly, the investigators found no significant change in corneal power with age for this cross-sectional investigation from infancy through the teenage years, raising the question of whether the increased corneal power observed in this population was a structural difference that pre-disposed individuals to an increased risk of keratoconus rather than being a manifestation of pre-clinical keratoconus.45

Marsack et al. proposed a similar hypothesis after performing an analysis of corneal topography measures obtained on 132 individuals with Down syndrome aged 8–55 years and 136 controls aged 7–59 years.46 In the study, keratoconus detection indices (Inferior–Superior steepening and KISA% which is an algorithm that topographically quantifies the phenotypic features of keratoconus47) were used to identify eyes meeting pre-established criteria for the detection of keratoconus. An emphasis of the study was the false detection that can occur if indices are applied to topographic images of poor quality; a common occurrence when imaging individuals with Down syndrome. Thus, images were screened by a masked examiner to ensure that only good-quality images were included in the analysis. In applying the detection indices, a greater percentage of eyes from individuals with Down syndrome were identified as suspects for keratoconus than found in the control population (Table 3). However, when comparing corneal power to the criteria for classification of keratoconus established by the Collaborative Longitudinal Evaluation of Keratoconus (CLEK) study,48 the majority of individuals with Down syndrome identified as suspect for keratoconus had corneal power in the moderate range (45–52 D) rather than the more even distribution of moderate and severe presentations observed in the CLEK study's population of individuals with known keratoconus (Table 4). Lastly, Marsack et al. calculated the difference in steep corneal power between fellow eyes of individuals with Down syndrome, both detected as suspicious for keratoconus and those not detected as suspicious, and found that inter-eye symmetry was more consistent with that observed in healthy myopic subjects than those with a known diagnosis of keratoconus (Table 5).

TABLE 3.

Percentage of eyes detected as suspect for keratoconus by two different detection indices applied to corneal topography measures.

	Eyes of subjects with Down syndrome (n = 221)	Eyes of control subjects (n = 274)
I–S > 1.4	46 (20.8%)	6 (2.2%)
KISA% > 100	26 (11.8%)	0 (0.0%)

Note: I-S, Difference between the inferior and superior corneal powers in dioptres and KISA% (an algorithm that topographically quantifies the phenotypic features of keratoconus). Data from Marsack et al.46

TABLE 4.

Distribution of mild, moderate and severe keratoconus classifications for individuals with Down syndrome identified with one or both eyes as suspect for keratoconus, compared with the distribution of participants in the Collaborative Longitudinal Evaluation of Keratoconus (CLEK) study with known keratoconus.

Steep Keratometric Reading	Individuals with Down syndrome with both eyes available for analysis—One or more eyes detected as suspect for keratoconus	CLEK keratoconus subjects
	% (n)	% (n)48
Mild (<45 D)	7.1 (2)	4.6 (55)
Moderate (45–52 D)	89.3 (25)	48.7 (586)
Severe (>52 D)	3.6 (1)	46.7 (562)

Note: Data from Marsack et al.46

TABLE 5.

Difference in steep corneal power (K) in dioptres (D) between the fellow eyes of individuals from the Collaborative Longitudinal Evaluation of Keratoconus (CLEK) study with and without keratoconus, compared with individuals with Down syndrome (DS) who were either detected as having an eye suspicious for keratoconus or not.

	CLEK myopic subjects49 (n = 330)	Individuals with Down syndrome with neither eye detected (n = 61)	DS subjects with one or more eyes detected (n = 27)	CLEK keratoconus subjects49 (n = 1068)
Steep K between Eyes (D)	−0.36 ± 0.37	−0.60 ± 0.85	−0.50 ± 0.40	−4.35 ± 4.41

Note: Data from Marsack et al.46

The question still remains whether corneal differences observed in individuals with Down syndrome are sequelae of pre-clinical keratoconus or rather stable structural differences that leave an individual at greater risk for keratoconus. The possibility of the association between Down syndrome and keratoconus is not surprising, given the previously described thinner and steeper corneas and elevated levels of refractive astigmatism in the eyes of individuals with Down syndrome,8,9,23 which are common findings with the eyes of individuals diagnosed with keratoconus. Dissimilar between groups, however, is that the refractive findings in individuals with Down syndrome tend towards more hyperopic refractive errors,9,15 whereas individuals diagnosed with keratoconus tend to exhibit elevated levels of myopia.49

Corneal topography is now ubiquitous in the management of keratoconus and in today's clinic, topographically derived metrics are used in conjunction with the more subjective, slit-lamp observations that are the current standard for the diagnosis of keratoconus. Detection metrics take many forms but share the design of distilling measured corneal morphology data into single-valued numerical indices that can be compared with threshold values determined from analysis of a known population of individuals with keratoconus. If the Down syndrome population has corneal morphology consistent with keratoconus, it is hypothesised that the thresholds for indices that have been found to detect keratoconus (in a population without Down syndrome) should also detect the presence of keratoconus-like morphology in the Down syndrome population. However, it remains unknown if the corneal structure observed in individuals with Down syndrome is prone to progressive deterioration, similar to that observed in keratoconic eyes of individuals without Down syndrome, and thus applying the same thresholds for detection to a population of individuals with Down syndrome may falsely identify individuals who do not progress to subsequent disease.

Wavefront aberrations

Given the differences in corneal structure observed in individuals with Down syndrome, it is reasonable to expect that greater optical distortions exist, as is seen in eyes with keratoconus. Two laboratories have reported higher-order wavefront aberration measures in individuals with Down syndrome. McCullough et al. obtained cycloplegic measures with the ImagineEyes Shack-Hartmann wavefront aberrometer (Imagine Eyes, imagine-eyes.com) in 30 children with Down syndrome aged 5–16 years, and compared root-mean-square (RMS) higher-order aberrations to 198 age-matched controls without Down syndrome.10 Their results found a statistically significant higher level of aberrations in the children with Down syndrome as compared with the controls (Table 6), but not to the pathological levels typically seen in conditions such as keratoconus.10 Anderson et al.50 reported similar findings for a cohort of 27 adults with Down syndrome measured with the Discovery Systems wavefront aberrometer (previously manufactured by Innovative Visual Systems, Elmhurst, llinois) post-dilation and compared with age-matched controls (Table 6), as well as a second cohort of 30 adults measured with the COAS wavefront aberrometer (previously manufactured by Johnson & Johnson Vision, Santa Ana, California) post-dilation.11 Both laboratories concluded that while the elevated aberrations observed in individuals with Down syndrome may contribute to the visual acuity deficits observed in this population, they were not elevated to the magnitude observed in individuals with corneal disease. A summary of higher-order RMS for controls, Down syndrome and keratoconus patients is shown in Table 6, organised by the pupil diameter through which the wavefront analysis was performed.

TABLE 6.

Published values of root-mean-square (RMS) higher-order aberrations for controls, individuals with Down syndrome and individuals with keratoconus.

	Study	3 mm pupil	4 mm pupil	5 mm pupil	6 mm pupil
Controls	McCullough et al.10 Control children	0.09 (IQR: 0.07—0.11)		0.28 (IQR: 0.22–0.34)
	Anderson et al.50 Control adults		0.13 ± 0.04		0.43 ± 0.17
	Salmon et al.51 Control adults	0.045 ± 0.021	0.100 ± 0.044	0.186 ± 0.078	0.327 ± 0.130
Down syndrome	McCullough et al.10 Children with Down syndrome	0.110 (IQR: 0.10–0.14)		0.34 (IQR: 0.26–0.41)
	Anderson et al.50 Adults with Down syndrome		0.21 ± 0.09		0.63 ± 0.34
	Anderson et al.11 Adults with Down syndrome		0.20 ± 0.07
Keratoconus	Pantanelli et al.52 Adults with Keratoconus				2.24 ± 1.22
	Kosaki et al.53 Adults with Keratoconus		0.72 ± 0.35

Note: In addition, the age range of the groups is indicated.

Abbreviations: IQR, Inter-quartile range.

Optical corrections for individuals with Down syndrome

The following studies were conducted to evaluate optical treatments for the deficits observed in individuals with Down syndrome with the goal of maximising visual acuity, either at distance or near. These included metric-optimised wavefront refraction strategies for distance corrections and the use of bifocal additions for near.

Utilisation of image quality metrics to identify refractions

The evidence for an optical limit to visual acuity is encouraging in that current treatment modalities, such as spectacles, can often compensate for a portion of these deficits. However, clinical techniques to determine the optimum spectacle prescriptions are currently under-serving the Down syndrome community. Individuals with Down syndrome are often unable to participate in the cognitively demanding aspects of subjective refraction, leaving clinicians to base prescriptions on objective clinical measurements of lower-order refractive error (i.e., sphere and cylinder). This is problematic, given that lower- and higher-order aberrations interact to impact retinal image quality,54–56 suggesting that a full sphero-cylindrical correction may exacerbate the effects of higher-order aberrations in more aberrated eyes, such as those of the Down syndrome population. Although higher-order aberrations cannot be corrected with standard spectacle lenses, if they are measured in a clinical setting, then their contribution to the resultant visual performance for a given spectacle correction can at least be considered. By utilising computer analysis, one could theoretically apply a spectacle correction to the measured optics of the eye and then predict the resultant image quality to determine whether it is an appropriate correction.57

Significant work has been done to enable the analysis of such measures for aberrated eyes by calculating the resultant image quality with various mathematical metrics. Thibos et al.58 described 31 different metrics that can be used to provide a single value of image quality. These metrics weigh different aspects of image quality (contrast, blur or doubling, for example) and thus may not all rate a particular optical outcome with the same value. However, several of these metrics have been demonstrated to correlate well with actual visual acuity measures,59 suggesting that they can likely be utilised as part of an objective refraction strategy for individuals with elevated higher-order aberrations. By calculating the resultant metric values for a variety of refractive corrections, results can be ranked and the refraction providing the best metric value identified as the one predicted to be the optimal correction for a given eye.57

Ravikumar et al. pursued this strategy for individuals with Down syndrome and sought to determine which of the 31 metrics would be most likely to identify refractions that maximised visual acuity for patients with Down syndrome. In addition, the team sought to determine whether metric-optimised refractions are predicted to out-perform the habitual corrections worn by individuals with Down syndrome.60 In this study, dilated wavefront measures were obtained with the COAS wavefront aberrometer (previously manufactured by Johnson & Johnson Vision, Santa Ana, California) for 30 adults with Down syndrome, and the habitual spectacle corrections were determined from lensometry. Next, custom software was run in MATLAB (mathworks.com) to apply >25,000 sphero-cylindrical combinations to the wavefront error of the eye and calculate the resultant value for 31 image quality metrics for each correction tested. Refractions were then ranked for each of the respective metrics and the top-rated refraction identified for each one.

Next, visual acuity charts were generated to simulate the retinal image quality in each eye for each top-ranked refraction by convolving a clear chart with the point spread function determined from the residual wavefront error for each condition using Image Simulation software (Sarver and Associates, saavision.com). Visual acuity charts simulating the retinal image quality for the participant's habitual refractive correction, an open-field autorefraction measure and a theoretical zeroing of the lower-order aberrations were also generated. As a first pass at reducing the number of metrics considered, 15 that were judged to consistently identify refractions that resulted in poor chart quality were eliminated. The remaining 16 metrics and the 3 additional refractive corrections (habitual, autorefraction and lower-order zeroing) were included in a chart reading experiment.

For the chart reading experiment, control observers without Down syndrome viewed simulated charts on a high-contrast liquid crystal display monitor through a unit magnification telescope with a 3 mm pupil aperture after dilation with 1% tropicamide and 2.5% phenylephrine. Controls were corrected with their habitual correction while performing the experiment. Charts were randomised and presented in blocks of 60–75 charts, and multiple experimental sessions were conducted. During each session, control observers would read the charts to determine visual acuity for each respective condition. Within each set of charts, a clear, unaberrated chart was presented to define the baseline visual acuity of the observer. Figure 10 shows an example of a simulated chart and the experimental setup for the control observer viewing the chart.

FIGURE 10.

Example of simulated chart representing a blurred image from elevated aberration levels (a) and experimental setup for control participant chart reading experiment (b). Figures previously published by Ravikumar et al. in 2019.60

The resultant visual acuity for all of the simulated charts was calculated as the difference between clear and simulated chart acuity and represented as ‘letters lost’ on a logMAR scale. Table 7 shows an example of the refractions identified for each of the conditions evaluated for a single eye from one adult with Down syndrome, and the average resultant number of letters lost for the five control observers reading those charts. The data in Table 7 demonstrate that for this particular eye, habitual correction provided the worst visual acuity, followed by autorefraction as second worst. Many of the metric-optimised refractions showed a multi-line improvement over the habitual correction.60

TABLE 7.

Refractions identified for the 16 metric optimisations and the three additional refractive conditions (habitual, autorefraction [autoref] and theoretical lower-order zeroing [LOAZ]) with corresponding average relative acuity for the five control observers.

Condition	Average relative acuity (logMAR)	Sphere (D)	Cylinder (D)	Axis (degrees)
HABITUAL	−0.54	+0.50	−1.75	20
AUTOREF	−0.37	−2.25	−1.50	30
NS	−0.24	−1.75	−1.25	40
SROTF	−0.20	−1.00	−2.25	42
PFCT	−0.19	−1.50	−1.25	42
LIB	−0.14	−1.75	−1.00	48
SRMTF	−0.14	−1.75	−1.00	48
SRX	−0.14	−1.75	−1.00	48
STD	−0.14	−1.75	−1.00	48
VSOTF	−0.14	−1.75	−1.00	48
PFST	−0.12	−1.5	−1.25	38
VSX	−0.12	−1.75	−1.25	44
AREAOTF	−0.11	−1.75	−1.00	50
LOAZ	−0.11
ENT	−0.11	−1.75	−1.25	36
PFCC	−0.10	−1.50	−1.50	44
PFWC	−0.10	−1.50	−1.50	44
VSMTF	−0.10	−1.75	−1.25	42
AREAMTF	−0.08	−1.75	−1.00	46

Note: The 16 metric optimisations listed here are neural sharpness (NS), Strehl ratio for optical transfer function (SROTF), pupil fraction for curvature tessellation (PFCT), light in the bucket (LIB), Strehl ratio for modulation transfer function (SRMTF), Strehl ratio in space domain (SRX), standard deviation of intensity (STD), visual Strehl ratio for optical transfer function (VSOTF), pupil fraction for slope tessellation (PFST), visual Strehl in space domain (VSX), area of visibility for optical transfer function (AREAOTF), entropy (ENT), pupil fraction for curvature critical pupil (PFCC), pupil fraction for wavefront critical pupil (PFWC), visual Strehl ratio for modulation transfer function (VSMTF) and area of visibility for modulation transfer function (AREAMTF). Data from Ravikumar et al.60

For each eye, visual acuity was considered in identifying the best-performing metric for that eye, and any metric identifying a refraction that yielded a visual acuity within 0.09 logMAR of the best-performing metric was subsequently identified as top performing. Figure 11 shows the percentage of eyes (n = 60) for which each condition was best performing and top performing. These data reveal that habitual corrections and autorefraction were among the lowest ranked across the group of eyes tested, suggesting that metric-optimised refraction may better serve this population. The overall visual acuity performance for habitual refractions versus best-performing metric refractions is shown in Figure 12, providing strong evidence that applying metric-optimised refractions to individuals with Down syndrome is predicted to outperform the refractions they were habitually wearing.60

FIGURE 11.

Distribution of best and top-performing metrics for all 60 eyes. Best performing indicates the refraction resulted in best average visual acuity as judged by control observers, whereas top performing means that the refraction resulted in an average control observer within 0.09 logMAR of the best visual acuity for a given patient's eye. The refraction methods shown here are optimisation of the following metrics: Standard deviation of intensity (STD), visual Strehl ratio for modulation transfer function (VSMTF), visual Strehl in space domain (VSX), neural sharpness (NS), Strehl ratio in space domain (SRX), light in the bucket (LIB), visual Strehl ratio for optical transfer function (VSOTF), area of visibility for modulation transfer function (AREAMTF), Strehl ratio for modulation transfer function (SRMTF), area of visibility for optical transfer function (AREAOTF), Strehl ratio for optical transfer function (SROTF), pupil fraction for curvature tessellation (PFCT), pupil fraction for wavefront critical pupil (PFWC), pupil fraction for curvature critical pupil (PFCC), entropy (ENT) and pupil fraction for slope tessellation (PFST). In addition, a refraction method representing a theoretical lower-order zeroing (LOAZ), an autorefraction (AUTOREF) and the participants presenting correction (HABITUAL) are included. Figure previously published by Ravikumar et al. in 2019.60 It should be noted that the abbreviation PFCTC in the figure should be PFCC and was previously published in error.

FIGURE 12.

The relative visual acuity measured from control observers with charts simulating the habitual corrections of eyes from participants with Down syndrome, as compared to the best-performing metric refraction. Figure previously published by Ravikumar et al. in 2019.60

Clinical trial of metric-optimised refractions

Utilising the findings of Ravikumar et al.,60 the metrics visual Strehl in the space domain (VSX) and pupil fraction for slope (tessellation) (PFSt) were pursued as the metrics to optimise in a randomised clinical trial of spectacle prescribing for adults with Down syndrome.11 VSX and PFSt were selected due to a previous investigation demonstrating that they each have a strong correlation with visual acuity,59 the fact that they frequently identified different refractions60 and that one includes the neural contrast sensitivity function in its calculation (VSX), whereas the other (PFSt) does not.58 The latter is relevant because the neural contrast sensitivity included in the calculation of VSX is not based on data from an individual with Down syndrome, and thus its application to this population makes an assumption that the neural contrast sensitivity functions between individuals with and without Down syndrome would be similar.

For their clinical trial, Anderson et al.11 enrolled 30 adults with Down syndrome between 18 and 52 years of age who were free from nystagmus and anisometropic or strabismic amblyopia (≥3 lines difference in visual acuity between the two eyes). Participants underwent a complete eye examination performed by an expert clinician with >35 years of experience providing care for individuals with intellectual disability. The expert clinician identified a clinical refraction for each participant by utilising common clinical techniques, such as autorefraction, retinoscopy, subjective refraction and additional measures of retinoscopy and autorefraction following instillation of 1% tropicamide and 2.5% phenylephrine. Dilated wavefront aberration measures were taken at the end of the complete eye examination and analysed post-visit to identify two different metric-optimised refractions for each of VSX and PFSt. The three spectacle prescriptions (clinical, VSX optimised and PFSt optimised) were then manufactured using identical spectacle frames and dispensed in random order for 2 months each. The 2-month adapted distance visual acuity measured with letter-by-letter scoring on a logMAR style chart was the primary outcome measure. Participants were also dispensed a temperature sensor data logging device mounted to the temple of each frame to monitor spectacle compliance throughout the course of the study.

Participants in the study were compliant with spectacle wear, irrespective of the method used to identify the prescription (p = 0.35). Spectacle wear ranged from an average of 2 h per day to as much as 18 h per day across participants, with an approximate group average of 11 h per day. For the group as a whole, there was no significant effect of prescription type on binocular distance visual acuity (p = 0.34), but there was a significant effect of prescription type on monocular distance visual acuity (p = 0.01), with VSX-optimised refractions resulting in a 0.04 logMAR improvement over the other two prescription types.11 Group mean estimates in logMAR visual acuity are shown in Table 8 for both binocular and monocular distance visual acuity for each prescription type. When comparing individual participants' visual acuity with each treatment to the presenting acuity at the first study visit, the majority of participants demonstrated improvements in visual acuity with at least one treatment type dispensed in the study (Figure 13).

TABLE 8.

Visual acuity outcomes after 2 months of adapted wear time for each prescription type.

Viewing condition	Treatment	Estimate (logMAR)	95% CI
Binocular visual acuity	PFSt	0.31	0.26–0.36
	VSX	0.33	0.27–0.38
	Clinical	0.34	0.28–0.39
Monocular visual acuity	PFSt	0.41	0.35–0.46
	VSX	0.37	0.31–0.42
	Clinical	0.41	0.35–0.47

Note: PFSt is the refraction determined from optimisation of the metric pupil fraction for slope tessellation. VSX is the refraction determined from optimisation of the metric visual Strehl in space domain. CI is the 95% confidence interval. Data from Anderson et al.61

FIGURE 13.

Individual participants change in visual acuity from the habitual presenting value with the three prescription types dispensed in the clinical trial. PFSt is the refraction determined from optimisation of the metric pupil fraction for slope tessellation. VSX is the refraction determined from optimisation of the metric visual Strehl in space domain. The change in visual acuity was calculated based on binocular distance values after 2 months of treatment wear. Note that negative values represent improved visual acuity with the treatment. Figure previously published by Anderson et al. in 2022.61

Although the visual acuity gains observed in the clinical trial were not as large as those predicted by the simulation study, both the expert clinician and metric-optimised refractions provided an improvement over the participants' habitual correction. An expert clinician was selected for the study to represent a high level of clinical care currently available for individuals with Down syndrome. However, it is suspected that many practitioners do not feel confident in caring for this community, or simply do not have the volume of experience from which to draw in their clinical decision-making. Although the metric-optimised refractions in this study were essentially equivalent to the clinician refractions, not all individuals with Down syndrome have access to a clinician expert, and thus the implementation of metric-optimised refraction in clinical practice may increase the quality of refractions received by individuals with Down syndrome even in clinical settings having minimal experience with this group. Additionally, the participants in the clinical trial were 18 years of age or older, and thus it is possible that greater gains in visual acuity were not observed due to long-standing bilateral amblyopia from sub-optimal spectacle corrections. Future work evaluating this prescribing technique in younger children with Down syndrome may yield greater visual acuity gains.

Bifocal spectacles

Given the almost universal presence of high accommodative lags in children with Down syndrome, several studies have investigated whether bifocals are beneficial for these individuals. The first reported controlled study was conducted by Stewart et al. on children with Down syndrome aged 5–11 years.62 Thirty-four children were randomly assigned to either the single vision or +2.50 add bifocal group. Accommodation was measured with dynamic retinoscopy at baseline and approximately 1, 7 and 19 weeks after dispensing new spectacles. Accommodation was analysed by calculating the accommodative error index,63 a single value that represents the accuracy of accommodation over the range of stimuli tested: In this study, demands of 4, 6 and 10 D. The authors found that for the final study visit, the accommodative error index improved by a mean of 0.45 ± 1.09 for the single vision control group, 2.88 ± 1.12 for the bifocal group when viewing through the bifocal and surprisingly by 1.66 ± 0.96 D for the bifocal group when viewing through the distance portion of the spectacles.62 These findings support the hypothesis that bifocals are effective in improving the near focus of patients with Down syndrome and suggest that they may not only be a passive treatment for close distance viewing but also an active treatment at near given the improvement in accommodation observed through the distance portion of the spectacles after approximately 5 months of bifocal wear.

A later study by the same laboratory compared accommodative responses at the time of bifocal prescription and with a longer follow-up time averaging 3.41 years (range = 1–7.8 years).64 Al-Bagdady et al. analysed accommodative responses through the distance correction for three stimulus demands (4, 6 and 10 D) using dynamic retinoscopy and found that 26 of 40 children showed an improvement in accommodation at the follow-up visit and 14 of those had accurate accommodation when compared to age norms of typical children65 after bifocal wear.64 At the time of the investigation, six individuals had discontinued bifocal wear and maintained accurate accommodation over a follow-up period ranging from 1.53 to 5.02 years. This supports the earlier findings62 that bifocals may be an active therapy and provides further evidence of retention in the improvement of accommodative function for children with Down syndrome who discontinued bifocal wear.

Nandakumar and Leat conducted a separate prospective study of 14 children aged 8–18 years with Down syndrome, of whom 11 were prescribed bifocals and evaluated both 2 weeks and 6 months after bifocals were dispensed.66 Bifocal power was customised for each participant based upon the accommodative lag at either 4 D or 6 D (whichever was closest to the participant's habitual working distance), resulting in a range of prescribed add powers from +1.00 to +3.50. The study found that near visual acuity improved with bifocals, both at the initial (p = 0.01) and 6-month visit (p = 0.02) when compared to near visual acuity with single vision lenses. The group mean improvement was approximately 1.5 lines as interpreted from the published figure. The study also found an improved accommodative response when tested through the bifocal at both 2 weeks (p = 0.02) and 6 months (p = 0.002) when compared with measures obtained through single vision lenses before the bifocals were dispensed. However, unlike the studies of Stewart et al.62 and Al-Bagdady et al.,64 Nandakumar and Leat reported no improvement in accommodative lag when measured through the distance portion of the bifocals at 6 months post bifocal wear. While this study did not support reports that bifocals may be a rehabilitative treatment to improve baseline accommodative function, it did show significant improvements in visual perceptual and literacy skills at the 6-month visit, which argues for a benefit of bifocals beyond simply improving near visual acuity.66

Perhaps the largest study of bifocals in individuals with Down syndrome was a multi-centre, randomised controlled trial of 119 children with Down syndrome aged 2–16 years conducted by de Weger et al.67 In this trial, children were randomised either to a full-cycloplegic refractive error correction in single vision lenses or to full-cycloplegic refractive error correction in a bifocal with a +2.50 addition. Measures of accommodation with dynamic retinoscopy with targets at 25 and 16.7 cm, as well as near visual acuity at 40 cm, were taken at 6 weeks, 6 months and 1 year after dispensing the corrections. A total of 102 children completed the final measurements. Time point analysis showed a significant improvement in near visual acuity for children randomised to the bifocal group compared with those wearing single vision lenses at the final study time point of 1 year (0.29 ± 0.20 logMAR vs. 0.42 ± 0.26 logMAR, p = 0.006).67 It should be noted that visual acuity measures were obtained through the bifocal addition for that group. Similar to Nandakumar et al.,66 this study found no significant change in accommodative accuracy through the distance correction across time.

Summary

In summary, individuals with Down syndrome have significant ocular conditions and visual deficits that introduce added barriers to the completion of activities of daily living and their ability to achieve independence. The findings of the work summarised here point to several sources, both optical and sensory, that may contribute to the reduced visual acuity commonly observed in this population. Recent work, including a clinical trial, has demonstrated promising new methods to optimise spectacle prescribing strategies that may enable broader successful refractive management of this population. In addition, the numerous studies demonstrating the benefit of prescribing bifocal lenses, whether as passive or active therapy, can be implemented immediately by all practitioners to achieve gains in near vision for both children and adults with Down syndrome. Much work remains to be done, and the restoration of normal visual acuity may not be possible without addressing the contribution of other sensory and neural deficits to visual performance, but there can be meaningful gains in vision quality for patients with Down syndrome now by educating practitioners on the advancements that have already been made.

AUTHOR CONTRIBUTIONS

Heather A. Anderson: Conceptualization (equal); data curation (equal); formal analysis (equal); funding acquisition (equal); investigation (equal); methodology (equal); project administration (equal); resources (equal); supervision (equal); validation (equal); visualization (equal); writing – original draft (equal); writing – review and editing (equal).

CONFLICT OF INTEREST STATEMENT

The authors of this manuscript has no conflicts of interest to disclose.