Ophthalmological outcomes, visual perception, fine motor precision, and visual‐motor integration in children born very preterm

Abstract

Aim

To explore longitudinal associations between ophthalmological outcome and development of visual perception, fine motor precision, and visual‐motor integration (VMI) in children born very preterm (gestational age <32 weeks).

Method

This was a prospective longitudinal cohort study of 113 children (51 females, 62 males) born very preterm. Visual acuity, low‐contrast visual acuity, stereopsis, and strabismus were investigated at 2 years 6 months (n = 98), 6 years 6 months (n = 84), and 12 years (n = 59). VMI was assessed at 6 years 6 months and 12 years, and visual perception and fine motor precision at 12 years, with the Beery‐Buktenica Developmental Test of Visual‐Motor Integration. Controls born at term were assessed at 6 years 6 months (n = 64) and 12 years (n = 46).

Results

For children born preterm, subnormal visual acuity at 2 years 6 months was related to both poorer VMI and fine motor precision at 12 years. Subnormal visual acuity, strabismus, and stereopsis at 6 years 6 months were also related to poorer VMI, visuo‐perceptual, and fine motor measures at 12 years. These associations were not present in controls born at term.

Interpretation

Subnormal ophthalmological function at 2 years 6 months and 6 years 6 months in children born very preterm is related to poorer performance on VMI, visuo‐perceptual, and fine motor precision tasks at 12 years, indicating that the quality of visual input interacts with developing visual‐motor abilities. The results of this study provide new insights on the relationship between ophthalmological function and neurodevelopment after preterm birth.

Children born very preterm with poorer ophthalmological function at 2 years 6 months, 6 years 6 months, and 12 years showed poorer visuo‐perceptual, visual‐motor, and fine motor skills at 12 years, emphasizing the role of visual input in visual‐motor development. These associations were not found in controls born at term, highlighting the unique risks faced by children born preterm. The findings provided new insights into how early ophthalmological deficits impact long‐term neurodevelopment after preterm birth.

What this paper adds.

• Ophthalmological function is related to motor and perceptual skills in children born very preterm.

• These associations are present from 2 years 6 months to 12 years in these children.

• Quality of visual input probably influences visual‐motor development in children born preterm.

• Relative scores for visual‐motor development decreased between 6 years 6 months and 12 years.

Abbreviations

Beery VMI

Beery‐Buktenica Developmental Test of Visual‐Motor Integration

CVI

cerebral visual impairment

IVH

intraventricular haemorrhage

ROP

retinopathy of prematurity

VMI

visual‐motor integration

Decreased ophthalmological function, affecting visual perception, visual‐motor integration (VMI), and visually guided activities, constitutes concern in children born very preterm, ¹ , ² with problems tending to persist into adulthood. ³ , ⁴ The abilities are central for general development and are associated with executive functions and academic achievement. ⁵ , ⁶ , ⁷ Visual functions and visually guided activities rely on widespread neural networks, which make them susceptible to perinatal insults. ⁸ , ⁹

Ophthalmological function, including visual acuity, contrast sensitivity, and stereopsis, provides the foundation for the quality of visual input. Visual perception organizes the visual input to meaningful information necessary for guiding and receiving feedback on motor activities. VMI, the ability to coordinate motor responses based on interpreted visual input, has a critical role in activity‐dependent development during early childhood. ¹⁰ Therefore, the development of visual‐perceptual and visual‐motor cortical networks may depend on the quality of visual input.

Retinopathy of prematurity (ROP) is a retinal vascular disorder affecting infants born preterm. Atypical development of retinal blood vessels potentially leads to retinal detachment and visual impairments. Severe ROP is a predictor of neurodevelopmental deficits. ¹¹

Cerebral visual impairment (CVI) is visual dysfunction resulting from damage to the visual pathways or cortical areas responsible for processing visual information. It consists of difficulties in interpreting visual input, together with or independent of, ophthalmological deficits. Furthermore, it can manifest as difficulties in VMI, visual perception, and visually guided fine motor skills. ¹¹ , ¹²

Previous reports of the LOngitudinal study of VISuomotor capacity (LOVIS) study group of children born very preterm showed impairments in ophthalmological outcomes at 2 years 6 months and 6 years 6 months, as well as deficits in VMI at 6 years 6 months and 12 years. ¹³ , ¹⁴

The aim of the present study was to investigate longitudinal associations between ophthalmological characteristics and results on the Beery‐Buktenica Developmental Test of Visual‐Motor Integration (Beery VMI) with supplemental tests of visual perception and motor precision in children born very preterm compared to controls born at term. Specifically, we investigated early (2 years 6 months and 6 years 6 months) or contemporary (12 years) ophthalmological function in relation to perinatal characteristics, VMI at 6 years 6 months and 12 years, and visual perception and fine motor precision at 12 years.

METHOD

Participants

Between 2004 and 2007, all live‐born infants with a gestational age below 32 weeks in Uppsala County, Sweden, were eligible for the population‐based LOVIS study. ¹⁴ Of 145 children, 113 were included (Figure S1). Neonatal data were recorded prospectively, ¹⁴ including gestational age, ROP, severe brain injury (periventricular leukomalacia and intraventricular haemorrhage [IVH], grade 3 or 4) (Table 1). Possible brain injury was assessed with cranial ultrasound during the first week after birth, at 35 to 36 gestational weeks, and close to term. ¹⁵ Screening for ROP was performed in the neonatal period and the criteria for treatment followed the ETROP study. ¹⁶ .

TABLE 1.

Study characteristics for the cohort born preterm.

Characteristic	LOVIS cohort, neonatal period, n = 113	LOVIS cohort, 2 years 6 months, n = 98	LOVIS cohort, 6 years 6 months, n = 84	LOVIS cohort, 12 years, n = 59
Female, n (%)	51 (45)	43 (44)	40 (44)	23 (39)
Gestational age, weeks, mean (SD)	28.7 (2.3)	28.8 (2.4)	28.6 (2.4)	28.6 (2.2)
Birthweight, grams, mean (SD)	1198 (353)	1235 (350)	1205 (346)	1213 (328)
IVH grade 3 or 4, n (%)	5 (4)	5 (5)	5 (6)	1 (2)
PVL, n (%)	7 (6)	7 (7)	7 (8)	4 (7)
Severe brain injury (PVL or IVH grade 3 or 4), n (%)	9 (7)	9 (9)	9 (11)	5 (8)
Retinopathy of prematurity any stage, n (%)	32 (28)	24 (25)	24 (28)	13 (22)
Retinopathy of prematurity ≥ stage 3, n (%)	9 (8)	8 (8)	4 (5)	2 (3)
Treated retinopathy of prematurity, n (%)	6 (5)	6 (6)	4 (4)	1 (2)
Cerebral palsy, n (%) a	NA	8 (8)	5 (6)	7 (12)*
Intellectual disability	NA	NA	NA	3 (5)
Cerebral visual impairment	NA	NA	NA	2 (3)
Visual impairment (visual acuity <0.3), n (%)	NA	1 (1)	1 (1)	1 (1)

Note: The cohorts born preterm at different ages were separately compared using a Student's t‐test for dimensional variables and a χ ² test for categorical variables at different times. Statistically significant differences between groups are denoted *p < 0.05.

Abbreviations: IVH, intraventricular haemorrhage; LOVIS, LOngitudinal study of VISuomotor capacity; NA, not applicable; PVL, periventricular leukomalacia.

^a Statistical significance refers to group differences between a cerebral palsy diagnosis at 2 years 6 months, 6 years 6 months, and 12 years.

Follow‐up assessments of neurodevelopment and visual function at the Departments of Paediatrics and Ophthalmology at Uppsala University Hospital occurred at 2 years 6 months (corrected for preterm birth), ¹³ 6 years 6 months, ¹⁴ , ¹⁷ and 12 years. ¹⁸ , ¹⁹ Neurodevelopmental diagnoses were obtained via parent report. At 6 years 6 months, a control group of age‐matched children born at term (n = 64, gestational age >37 weeks) was included, of which 46 also participated at 12 years. No control child had intellectual disability, cerebral palsy (CP), or CVI diagnoses. The Uppsala Regional Ethics Committee approved the study (Ups 03–665, 2016/400). All legal guardians signed the informed consent required to participate in the study.

Study protocol

Procedure

Ophthalmological examinations

The ophthalmological protocols for the assessments at 2 years 6 months, 6 years 6 months, and 12 years have been reported previously. ¹³ , ¹⁷ , ¹⁹ , ²⁰ Examinations were performed by an experienced orthoptist or paediatric ophthalmologist. Best‐corrected visual acuity at distance, that is, the best visual acuity taking any refractive error into account, was assessed monocularly and binocularly with logMAR Lea single optotypes at 2 years 6 months, with Lea linear optotypes at 6 years 6 months, and with logMAR Sloan letters at 12 years. Binocular near visual acuity was tested at 6 years 6 months and 12 years. Visual fixation and following a small moving torch were assessed if the child failed optotype testing. At 2 years 6 months, a visual acuity of less than 0.4, and at 6 years 6 months and 12 years a visual acuity of 0.8 or less Snellen decimal acuity, were considered subnormal. A cover test was used to detect strabismus. Stereo acuity was assessed at near with Lang stereo test at 2 years 6 months together with the TNO (The Netherlands Organisation for Applied Scientific Research) stereo test at 6 years 6 months and 12 years. TNO ≤60 arcsec was considered normal. Low‐contrast visual acuity was measured binocularly at 6 years 6 months and 12 years using the Lea Hyvärinen 2.5% low‐contrast test; results below 0.4 and 0.5 respectively were considered abnormal.

Test of visual‐motor integration, visual perception, and fine motor precision

The Beery VMI examined VMI, visual perception, and fine motor precision. The child was first asked to copy geometric shapes using paper and pencil (VMI subtest), then to match identical geometrical shapes according to parameters such as slope, spatial relations, and small details (visual perception subtest), and lastly to trace the shapes within specified boundaries (motor coordination subtest, here referred to as fine motor precision). ⁶ The subtests use the same 24 geometric shapes and are appropriate from 2 years to adulthood. ⁶ To make the visual perception subtest into an as pure visual‐perceptual task as possible, motor requirements were reduced by letting children point at the correct shapes. ⁶ In the motor precision subtest, visual‐perceptual and integrative demands were reduced by providing tracing paths and dots indicating where to start tracing with the pen. ⁶ Raw scores were transformed to normative scores (mean [SD] 100 [15]). The fifth and sixth editions of the Beery VMI were used for the 6 years 6 months and 12‐year assessments respectively. ⁶ , ²¹ The sixth edition contains a norm update. Therefore, differences between norms were checked using univariate regressions.

The Beery VMI was administered by licensed psychologists according to standardized administration guidelines. At 6 years 6 months, the children completed the VMI subtest only; at 12 years, they completed all three subtests. Results have been partially previously published. ¹⁴ , ²⁰ The present study only included results from children participating in the ophthalmological examinations, who all had Beery VMI results (Figure S1).

Statistical considerations

The results from the group born preterm and the control group were compared using a Student's t‐test for continuous variables and χ ² or Fisher's exact test for categorical variables; p < 0.05 was considered statistically significant. Univariate linear regressions explored associations between neonatal variables and outcome, and between ophthalmological characteristics and VMI, visual perception, and fine motor precision. Neonatal data and neurodevelopmental outcomes significantly associated with ophthalmological and Beery VMI data at 2 years 6 months, 6 years 6 months, or 12 years, were entered as covariates (i.e. were considered as confounders) in multivariate linear regressions with VMI at 6 years 6 months and 12 years; visual perception and fine motor precision performance at 12 years was the dependent variable. SPSS v28 (IBM Corp., Armonk, NY, USA) was used. No imputation was made for dropout or missing values. To control multiple comparisons in the regression analyses, the Benjamini–Hochberg procedure was applied with a conservative approach to the full set of univariate and multivariate regressions using a false discovery rate of 0.05. ²² Based on this correction, results with p ≤ 0.013 were considered statistically significant.

RESULTS

Table 1 presents the background characteristics and dropout analysis. Apart from a higher proportion of participating children with CP at 12 years than at 2 years 6 months, no differences were found. A comparison between children only attending the 12‐year neuropsychological follow‐up (n = 19) and the children participating in both assessments at 12 years (n = 59), showed that IVH grade 3 or 4 was less prevalent in those participating in both assessments (p = 0.047). No other neonatal characteristic, Beery VMI, or ophthalmological results at 6 years 6 months differed between the study and dropout groups.

Ophthalmological outcome at 12 years

Table 2 presents the ophthalmological outcome. Prevalence of low visual acuity, strabismus, and poorer stereo acuity was higher in the group born preterm than the control group. Gestational age and severe brain injury, any ROP, and ROP of 3 or greater were significantly associated with ophthalmological outcomes at 12 years in the univariate analyses (Table 3).

TABLE 2.

Ophthalmological study outcome at 12 years for the cohort born preterm compared with the control group born at term.

Ophthalmological outcome at 12 years	LOVIS cohort, n = 59	Control group born at term, n = 46
Distance binocular visual acuity, mean (SD)	1.16 (0.25)	1.19 (0.19)
Near binocular visual acuity, mean (SD)	0.96 (0.11)*	1.0 (.0.01)
Distance visual acuity ≤0.8, n (%)	8 (14)*	1 (2)
Distance visual acuity <0.3, n (%)	1 (2)	0
Low‐contrast visual acuity, mean (SD)	0.53 (0.17)	0.58 (0.12)
Subnormal contrast visual acuity <0.5, n (%)	15 (25)	5 (11)
Impaired stereo acuity, n (%)	16 (27)**	2 (4)
Impaired stereo acuity, strabismus excluded, n (%)	8 (16)	2 (4)
Any strabismus yes/no, n (%)	8 (14)**	0
Strabismus for distance, yes/no, n (%)	7 (12)*	0
Strabismus for near, yes/no, n (%)	8 (14)**	0

Note: The cohort born preterm was compared with the group born at term using a Student's t‐test for dimensional variables and a χ ² test for categorical variables. Statistically significant differences between the group born preterm and the controls born at term are denoted **p < 0.01 and *p < 0.05.

Abbreviation: LOVIS, LOngitudinal study of VISuomotor capacity.

TABLE 3.

Univariate regression between neonatal characteristics, neurodevelopmental outcome at 12 years, and Beery VMI score at 6 years 6 months and 12 years.

Variable		Gestational age	Any ROP	ROP ≥ stage 3	Severe brain injury	VMI 6 years 6 months	VMI 12 years	Visual perception 12 years	Fine motor precision 12 years
Any ROP	β	−0.668
	p	<0.001
ROP ≥ stage 3	β	−0.536	0.468
	p	<0.001	<0.001
Severe brain injury a	β	−0.043	0.050	0.012
	p	0.653	0.603	0.903
Visual‐motor integration 6 years 6 months	β	0.254	−0.187	−0.339	−0.048
	p	0.018	0.085	0.001	0.664
Visual‐motor integration 12 years	β	0.240	−0.146	−0.332	0.010	0.518
	p	0.034	0.203	0.003	0.931	<0.001
Visual perception 12 years	β	0.278	−0.160	−0.409	−0.210	0.447	0.443
	p	0.014	0.163	<0.001	0.065	<0.001	<0.001
Fine motor precision 12 years	β	0.215	−0.111	−0.237	−0.220	0.533	0.579	0.437
	p	0.059	0.334	0.037	0.053	<0.001	<0.001	<0.001
Cerebral palsy	β	0.001	−0.040	0.034	0.477	−0.090	−0.124	−0.170	−0.012
	p	0.989	0.676	0.791	<0.001	0.410	0.274	0.134	0.914
Cerebral visual impairment	β	−0.217	0.141	0.155	0.268	−0.227	−0.398	−0.386	−0.492
	p	0.021	0.138	−102	0.004	0.036	<0.001	<0.001	<0.001
Intellectual disability b	β	−0.174	0.141	0.358	0.111	−0.333	−0.357	−0.482	−0.350
	p	0.066	0.138	<0.001	0.242	0.002	<0.001	<0.001	0.002

Note: p‐values shown in bold represent statistically significant associations after corrections for multiple comparisons (p ≤ 0.013).

Abbreviations: Beery VMI, Beery‐Buktenica Developmental Test of Visual‐Motor Integration; ROP, retinopathy of prematurity; VMI, visual‐motor integration.

^a Either/or periventricular leukomalacia and intraventricular haemorrhage stage 3 or 4.

^b Clinical diagnosis of intellectual disability, that is, Full‐scale IQ less than −2SD as well as impaired everyday adaptive functioning.

Beery VMI scores at 6 years 6 months and 12 years

The two Beery VMI editions were highly correlated (β = 0.996, p < 0.001). We chose to use the original fifth edition norms for the VMI score analyses at 6 years 6 months. Figure 1 shows mean performance on the Beery VMI. For the children who participated at both 6 years 6 months and 12 years, VMI performance relative to the normative data deteriorated in the cohort born preterm with on average − 17.6 points compared to −9.1 points in the control group (p = 0.008). Figure 2 explores change at the individual level. In the univariate regression analyses, gestational age, severe brain injury, or an ROP of 3 or greater, were the most important neonatal factors associated with Beery VMI scores at 12 years (Table 3).

FIGURE 1.

Beery‐Buktenica Developmental Test of Visual‐Motor Integration (Beery VMI) performance at 6 years 6 months and 12 years compared for the group born very preterm and control group born at term. Visual‐motor integration (VMI) scores at 6 years 6 months for the group born preterm was a mean (SD) of 94.5 (15.6) and differed significantly from the controls (101.9 [12.9], p = 0.016). At 12 years, all three Beery VMI subscale scores differed between the groups. The mean (SD) scores for the VMI subtest for the group born preterm and the controls were 79 (15) versus 93 (12), p < 0.001, 95 (11) versus 102 (8), p = 0.003 for the visual perception subtest, and 86 (16) versus 98 (12), p < 0.001 for the fine motor precision subtest.

FIGURE 2.

Individual performance in the Beery‐Buktenica Developmental Test of Visual‐Motor Integration (Beery VMI) test at 6 years 6 months and 12 years for the group born very preterm and the control group. The two y‐axes on the left show the Beery VMI scores at 6 years 6 months; on the axes, the relationship to norms is shown in green (average, above −1SD), yellow (−1SD to −2SD), and red (below −2SD). The colours on the lines have the same categorization for the 12‐year Beery VMI assessment. Please note that the normative data for the Beery VMI was a mean (SD) of 100 (15); both groups showed poorer mean performance at the 12‐year VMI than at 6 years 6 months. This was more pronounced in the group born very preterm. In the group born preterm, 9 of 51 (18%) had changed from above −2SD at 6 years 6 months to below −2SD at 12 years compared to 1 of 46 (2%) in the control group.

Ophthalmological outcome in relation to the Beery VMI results

In the cohort born preterm, subnormal visual acuity at distance at 2 years 6 months was related to poorer 12‐year VMI and lower fine motor precision scores (β = 0.250, p = 0.031 and β = 0.232, p = 0.042). Table 4 presents the associations between the 6 years 6 months and 12‐year ophthalmological results and Beery VMI assessments. In the control group, only low‐contrast visual acuity at 6 years 6 months was related to the 12‐year visual perception scores.

TABLE 4.

Univariate and multivariate regression models between ophthalmological outcome in relation to Beery VMI performance for the cohort born very preterm.

	VMI 6 years 6 months, n = 71						VMI 12 years, n = 68						Visual perception 12 years, n = 69						Fine motor precision 12 years, n = 69
	Unadjusted		Adjusted for severe brain injury a and ROP ≥3		Adjusted for CVI and ID		Unadjusted		Adjusted for severe brain injury a and ROP ≥3		Adjusted for CVI and ID		Unadjusted		Adjusted for severe brain injury a and ROP ≥3		Adjusted for CVI and ID		Unadjusted		Adjusted for severe brain injury a and ROP ≥3		Adjusted for CVI and ID
6 years 6 months
	β	p	β	p	β	p	β	p	β	p	β	p	β	p	β	p	β	p	β	p	β	p	β	p
Distance binocular visual acuity	0.25	0.038	0.13	0.355	0.062	0.655	0.45	< 0.001	0.38	0.004	0.24	0.043	0.54	< 0.001	0.42	< 0.001	0.34	0.002	0.25	0.033	0.17	0.211	0.03	0.808
Near binocular visual acuity	0.38	0.001	0.30	0.020	0.249	0.063	0.46	< 0.001	0.46	< 0.001	0.26	0.037	0.49	< 0.001	0.34	0.004	0.26	0.027	0.45	< 0.001	0.42	0.002	0.26	0.033
Low‐contrast visual acuity	0.27	0.022	0.24	0.067	0.219	0.081	0.28	0.020	0.34	0.030	0.20	0.076	0.33	0.005	0.26	0.049	0.24	0.032	0.12	0.339	0.06	0.646	0.025	0.821
Subnormal contrast visual acuity <0.4	−0.07	0.543	−0.02	0.854	−0.036	0.769	−0.11	0.348	−0.10	0.421	−0.08	0.458	−0.18	0.141	−0.09	0.467	−0.15	0.175	0.39	0.003	0.12	0.338	0.09	0.384
Any strabismus	−0.39	< 0.001	−0.31	0.006	−0.032	0.004	−0.43	0.001	−0.34	0.004	−0.27	0.011	−0.24	0.051	−0.09	0.434	−0.12	0.233	−0.37	0.001	−0.28	0.021	−0.24	0.020
Stereopsis	−0.27	0.038	−0.20	0.108	−0.069	0.578	−0.27	0.045	−0.40	0.031	−0.10	0.393	−0.28	0.023	−0.23	0.052	−0.07	0.524	−0.28	0.016	−0.52	< 0.001	−0.11	0.346
12 years							VMI 12 years, n = 59						Visual perception 12 years, n = 58						Fine motor precision 12 years, n = 58
	β	p	β	p	β	p	β	p	β	p	β	p	β	p	β	p	β	p	β	p	β	p	β	p
Distance binocular visual acuity	NA		NA		NA		0.43	< 0.001	0.45	0.007	0.30	0.022	0.49	< 0.001	0.35	0.012	0.32	0.011	0.36	0.006	0.36	0.032	0.14	0.264
Near binocular visual acuity							0.45	< 0.001	0.47	0.002	0.33	0.022	0.42	< 0.001	0.27	0.040	0.19	0.189	0.45	< 0.001	0.46	0.001	0.02	0.104
Low‐contrast visual acuity							0.37	0.003	0.36	0.024	0.24	0.053	0.36	0.007	0.18	0.129	0.18	0.132	0.23	0.084	0.16	0.229	0.05	0.712
Subnormal contrast visual acuity <0.5							−0.48	< 0.001	−0.45	0.002	−0.38	0.001	−0.38	0.004	−0.23	0.051	−0.27	0.018	−0.16	0.234	−0.09	0.513	−0.04	0.711
Any strabismus							−0.48	< 0.001	−0.48	< 0.001	−0.35	0.013	−0.44	< 0.001	−0.28	0.026	−0.23	0.107	−0.54	< 0.001	−0.54	< 0.001	−0.36	0.008
Stereopsis							−0.39	0.002	−0.34	0.012	−0.25	0.048	−0.34	0.010	−0.24	0.082	−0.20	0.103	−0.26	0.045	−0.46	0.001	−0.31	0.012

Note: Covariates in the multivariate regressions are severe brain injury and retinopathy of prematurity stage 3 and above, as well as incidence of cerebral visual impairment and intellectual disability respectively.

p‐values marked in bold represent statistically significant associations after corrections for multiple comparisons (p ≤ 0.013).

Abbreviations: Beery VMI, Beery‐Buktenica Developmental Test of Visual‐Motor Integration; CVI, cerebral visual impairment; ID, intellectual disability; ROP, retinopathy of prematurity; VMI, visual‐motor integration.

^a Either/or periventricular leukomalacia and intraventricular haemorrhage stage 3 or 4.

Background factors and ophthalmological outcome in relation to the Beery VMI results at 12 years

Multivariable analyses revealed that severe brain injury and an ROP of 3 or greater were related to poorer 12‐year Beery VMI scores. Ophthalmological outcome at 6 years 6 months, but not 2 years 6 months, explained additional variance in the Beery VMI results (Table 4). The relationship between Beery VMI and ophthalmological measures at 12 years showed the same pattern of relations as at 6 years 6 months. In the multivariate regression, severe brain injury and an ROP of 3 or greater were entered as covariates; the effect of an ROP of 3 or greater disappeared but severe brain injury still explained unique variance in visual perception scores. In separate multivariate analyses, CVI and intellectual disability served as covariates and the results showed a continued relationship between ophthalmological outcome and the Beery VMI.

DISCUSSION

This study explored the links between ophthalmological outcomes and VMI, visual‐perceptual, and fine motor precision skills in children born very preterm. To our knowledge, this is the first study to investigate these links longitudinally in these children. The group born very preterm displayed deficits in both ophthalmological and Beery VMI measures compared to peers born at term. Subnormal visual acuity at 2 years 6 months was associated with poorer VMI and fine motor precision at 12 years. Early deficits in visual input may negatively affect long‐term development of the visual‐perceptual and visual‐motor integrative neural networks.

The abilities measured by the three Beery VMI subtests are linked because the subtests measure skills related to VMI. ⁶ We found moderate correlations between the subtests at 12 years, indicating that the three tests measure aspects of, but not entirely the same, abilities within VMI. ⁶ This was underscored by the difference in performance between the subtests, where VMI rendered the poorest scores compared to visual perception and fine motor precision in both groups.

Previous studies did not find contemporaneous associations between visual acuity and VMI in children ²³ or young adults born preterm. ³ These cohorts were born before 1991 and were not comparable to our cohort regarding gestational age or perinatal management. We found relationships between Beery VMI scores, visual acuity, and several other ophthalmological parameters, both longitudinally and contemporaneously. No relationships were confirmed in the controls, possibly because of low statistical power. The relationship between ophthalmological parameters and VMI, visual‐perceptual, and fine motor precision skills might also be unique in children born very preterm.

In typically developing children and adults, experimentally induced blurred vision (visual acuity at near) was associated with poorer visual‐motor abilities, indicating that poorer results were linked to vision rather than poor VMI ability. ²⁴ In our study, children were assessed with best‐corrected vision. The results showed longitudinal associations between poorer near visual acuity and poorer performance on all Beery VMI subtests for the group born preterm. Together, this suggests that the level of near visual acuity may influence VMI development over time.

Furthermore, in the present study, contrast visual acuity was linked to VMI and visual perception in the group born preterm at 12 years, but not at 6 years 6 months. Perhaps, the importance of low‐contrast visual acuity increases as the complexity of the Beery VMI figures rises because contrast visual acuity is influenced by optical conditions and retinal and neural processing. ²

Brain injury, such as IVH and periventricular leukomalacia, might cause reduced stereoacuity; ²⁵ abnormal stereopsis has been associated with poor performance on perceptual tests. ²³ In our study, severe brain injury and severe ROP were significant factors for lower Beery VMI scores. Subnormal stereopsis, even in the absence of strabismus, and lower visual acuity, were common in the group born very preterm and was related to VMI at 6 years 6 months and 12 years. When controlling for brain injury and ROP, the associations remained; thus, subnormal stereopsis explained unique variance for visual‐perceptual and fine motor precision.

Intellectual disability and CVI are diagnoses associated with visual‐motor difficulties; ¹ , ¹² , ²¹ in our study, they were related to poorer Beery VMI scores. The results showed the continued importance of ophthalmological outcome in relation to Beery VMI tasks, although attenuated. This was particularly seen in children with strabismus and poor stereopsis, and indicated that the relationship also persisted for children without these diagnoses. Cerebral palsy is known to be associated with VMI deficits but this was not mirrored in our results. ²⁶ The children with CP were all ambulatory and it is possible that their relatively mild motor disorder attenuated these relationships.

Structural changes in brain magnetic resonance imaging and visual‐perceptual dysfunction have been associated with ROP. ²⁷ , ²⁸ Our previous research indicated that severe ROP had a similar effect on cognition at 6 years 6 months and 12 years in this cohort. ¹⁴ , ¹⁸ Structural and functional abnormalities of the retina and reduced vision can also occur in the absence of ROP. ¹¹ , ²⁹ Cerebral lesions such as IVH and periventricular leukomalacia may cause injuries to the posterior visual pathway causing CVI. ⁹ , ¹² Exploring this notion, the hypothesis of ‘visuopathy of prematurity’ proposes a broader condition, encompassing both retinal vascular issues and visual processing abnormalities in children born preterm, either with or without ROP. ¹¹ The interplay between retinal abnormalities and cerebral lesions underscores the complexity of CVI in children born very preterm. Recognizing and addressing CVI is vital for optimizing developmental trajectories and enhancing quality of life for affected children. During the study period, a CVI diagnosis was reserved for children with visual‐perceptual difficulties and brain abnormalities. According to the hypothesis of visuopathy of prematurity, our results can be interpreted as supporting the use of a functional classification with less emphasis on detectable brain abnormalities. In accordance, we found that poorer visual‐motor and visual‐perceptual ability in children born very preterm were linked to both ROP and brain injury. However, as VMI challenges were also prevalent in children born very preterm without ROP or brain injury, VMI development probably has a multifactorial background. Our results imply that ophthalmological and perceptual development mature in synergy through childhood, a time period when plasticity of the visual system is pronounced.

Sensorimotor processes and activity‐dependent development are considered central driving forces for general development. ³⁰ Approximately 25% variance in VMI performance is explained by general intelligence. ¹ , ⁶ The importance of VMI is further underscored by the relationship between Beery VMI scores and academic achievement even when controlling for IQ. ⁷ Deficits in visual perception, VMI, and fine motor precision in children born very preterm are in accordance with previous studies. ¹ , ¹⁰ , ²⁰ , ²³ In this study, the quality of visual input was relevant for the development of higher‐order processing networks. Ophthalmological examinations for early detection and intervention of adverse ophthalmological outcomes are thus crucial in children born very preterm because the aggravated risks of preterm birth related to the visual system may alter the basic input on which further visual‐perceptual development relies. ¹⁰ This study represents a first exploratory step towards understanding if ophthalmological outcomes may be used to predict visual‐motor abilities at later ages. Potential differences in aetiology and developmental trajectories between children born very preterm with poor and normal ophthalmological outcomes remains to be explored.

Both the group born preterm and the control group showed surprisingly reduced rates of progress in VMI performance between 6 years 6 months and 12 years, which was more pronounced in the group born very preterm. This lack of expected progress was obvious in relation to the normative data included in the Beery VMI. Rescoring all results to the norms of the sixth edition did not alter the results. The finding is beyond the scope of this paper, but needs to be explored elsewhere.

In the present longitudinal study, differences in ophthalmological outcomes between children born very preterm and children born at term at 2 years 6 months and 6 years 6 months ¹³ , ¹⁷ remained at 12 years. The group born very preterm exhibited a higher prevalence of ophthalmological problems, including subnormal visual acuity at distance and near, strabismus, and impaired stereo acuity compared to controls. These results are in accordance with other longitudinal follow‐up studies. ² , ⁴ Neonatal characteristics, notably lower gestational age, severe brain injury, any ROP and ROP of 3 or greater, were all related to poorer ophthalmological outcomes.

Strengths and limitations

Limitations of this study included the sample sizes for the ophthalmological examination of the original cohort (follow‐up rate = 54%). An a priori sample size calculation was not performed because no previous work in this field has provided an empirical basis for estimating effect sizes. Future studies with larger samples and formal power analyses are needed to confirm and extend the study findings. Neonatal diagnosis of severe brain injury was used as a predictor, but a significantly lower proportion of children with IVH attended the examination, potentially hiding the effects of brain abnormalities. Both IVH and periventricular leukomalacia were assessed with repeated ultrasounds in the neonatal period, thus possibly failing to diagnose more subtle early brain injuries. Furthermore, children with intellectual disability, CVI, and CP were clinically diagnosed outside the study protocol, which may have introduced variability in case ascertainment. A strength of the study included the prospective longitudinal design for both ophthalmological and visual‐motor assessments.

Conclusions

Ophthalmological outcomes, both at preschool age and in early adolescence, were associated with visual perception, fine motor precision, and visual‐motor deficits in children born very preterm, independent of gestational age at birth. The same links between ophthalmological outcome and visual‐motor deficits were not found for controls born at term. In children born very preterm, deficits in ophthalmological outcome persisted into early adolescence. Screening for and treatment of ophthalmological deficiencies in children born very preterm should be undertaken because deficiencies in these functions are related to later visual perception, visual‐motor, and fine motor development, which in turn are related to general development and academic achievement.

Supporting information

Figure S1: Flow chart of the study population at the ophthalmological examinations.

Data S1: supporting Information.

DATA AVAILABILITY STATEMENT

The data set of the current study is not publicly available because of privacy reasons but it is available upon reasonable request to the authors.