INTRODUCTION

Alcohol is a well-known teratogen.^{1 2} Prenatal alcohol exposure might lead to a clinical diagnosis under the umbrella term of foetal alcohol spectrum disorders (FASD),³ which includes foetal alcohol syndrome (FAS), partial foetal alcohol syndrome (PFAS), alcohol-related birth defects and alcohol-related neurodevelopmental disorder (ARND).⁴ The most severe diagnosis under the FASD umbrella is FAS, and individuals with fully developed FAS have been shown to have ophthalmological abnormalities such as a short palpebral fissure, ptosis, strabismus, refractive errors and optic nerve hypoplasia (ONH).^5–8 Despite the high global incidence of FASD,⁹ surprisingly few studies on the effects of alcohol on the eye and visual system have been performed.

ONH has been considered a distinguishing feature in corroborating FAS and is traditionally diagnosed by an ophthalmologist using ophthalmoscopy. A small-sized optic disc can indicate ONH, and the presence of a double-ring sign and measurements taken from a fundus photography can strengthen the diagnosis.¹⁰ Severe cases of ONH are evident when using ophthalmoscopy, but subtle cases are sometimes not suspected until a pathological visual field test has been performed.¹¹ Static perimetry is sometimes challenging to use when examining children, but can also be difficult in adults who cannot co-operate perfectly, for example, due to attention deficits. On the other hand, most adults and many children over the age of 4 can co-operate with an optical coherence tomography (OCT) examination, which allows measurement of the size of the optic disc and the different layers of the retina, including both the macular and the peripapillary retinal nerve fibre layer (RNFL).The purpose of this study was to examine whether there were any differences between individuals with FASD and healthy controls in macular and peripapillary RNFL, retinal thickness and optic disc area (ODA) measured by OCT.

RESULTS

Demographic and ophthalmological data from the clinical examinations are summarised in table 1.

Of the 31 young adults with FASD who agreed to participate, one female participant discontinued the study, and another female participant was excluded from analysis of the OCT data due to a high degree of myopia (−12 D SE in the right eye and −13 D SE in the left eye). In three male participants, usable OCT images could not be obtained from either eye due to unsatisfactory fixation or frequent blinking during the scan. A further four participants had aggravating factors (ptosis, corneal decompensation, nystagmus and glare), leaving 22 participants with high-quality peripapillary OCT measurements of both eyes. Macular OCT measurements were available in 26 individuals, two of whom contributed with either the right eye or the left eye. Three controls were excluded from OCT analysis due to a high degree of myopia (n=2) and prematurity (n=1).

Optic disc size and configuration measured by OCT as well as the variables from the measurements of the macula are summarised in table 2.

Figure 1 shows the result of the OCT examination of the peripapillary RNFL, where the total thickness was significantly lower in the FASD group than in the controls. The total peripapillary RNFL was correlated to ODA in both eyes in the controls (right eye: r=0.47, p=0.013, left eye: r=0.43, p=0.03) and in one eye in the FASD group (right eye: r=0.37, p=0.089, not significant, left eye: r=0.53, p=0.007). Figure 2 describes the macular RNFL in nine ETDRS areas.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Peripapillary retinal nerve fibre layer (RNFL) of the right and left eye in 25 (22 contributed with both eyes) individuals with foetal alcohol spectrum disorders (FASD) and 27 healthy controls.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Macular retinal nerve fibre layer of the right and left eye in the nine ETDRS areas in 26 individuals (24 contributed with both eyes) with foetal alcohol spectrum disorders (FASD) and 27 healthy controls.

Four individuals in the FASD group, all of whom were diagnosed with FAS, were clinically detected with ONH (table 1 and figure 3). In one participant with ONH, we were unable to obtain OCT images of high quality. We also performed HFA perimetry in 10 individuals with FASD, but many participants had difficulties with the sustained attention and vigilance required by the test, and so the results were judged unreliable and excluded from analysis.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Ophthalmological data on four participants with clinically detected optic nerve hypoplasia, all of whom diagnosed with foetal alcohol syndrome.

A subanalysis of the FASD group comparing those with FAS (n=18) to those with PFAS or ARND (n=8) did not reveal any significant differences in ODA or thickness of peripapillary or macular RNFL. In the FASD group, total peripapillary RNFL was correlated to the SE in both the right eye (r=0.46, p=0.013) and the left eye (r=0.47, p=0.024). The more myopic the participant, the thinner the RNFL. Neither peripapillary nor macular RNFL was correlated to BCVA. A correlation between IQ and total RNFL was found to be moderate significant in right eye (r=0.49, p=0.026), but not in the left (r=0.29, p=0.17).

DISCUSSION

The results from this study showed lower retinal thickness, thinner RNFL in the outer areas of the macular RNFL, thinner total peripapillary RNFL, and lower BCVA in young adults diagnosed with FASD compared with healthy controls. No differences in ODA or refraction were found. This agrees with, to our knowledge, the only previous study measuring peripapillary RNFL in FASD, in which Menezes and colleagues found significantly thinner peripapillary RNFL both in total and in inferior and nasal quadrants among those with FAS (n=11) compared with controls.¹⁹ Their study only included participants with FAS, and not PFAS or ARND. However, we found significant differences in superior and temporal areas, they did not. This could be due to type 1 error, or the fact that different OCT machines were used. Further, they only measured the peripapillary RNFL, whereas we also measured macular RNFL and retinal thickness. We found that both peripapillary RNFL and the macular RNFL were thinner in most areas in individuals with FASD compared with healthy controls.

Budenz et al reported the mean peripapillary RNFL in a seminal large cross-sectional observational study of healthy individuals aged 18–85 years. The median peripapillary RNFL (105 μm in the right eye and 103 μm in the left eye) of our control group was similar to the mean peripapillary RNFL (103.7 μm) of their participants aged 18–29 years. In contrast, our participants with FASD had a median peripapillary RNFL (96.5 in the right eye and 96 μm in the left eye) that was equivalent to the mean RNFL (94.1 μm) of their participants aged 70–85 years.²⁰

Retinal thickness was generally lower in the FASD group, except for the central part of the retina, which was thicker although not statistically significantly. This may be because at least 10 (in three cases the gestational age was unknown) of our 26 participants were born preterm (gestational week 31–36), and a thicker central retina has been associated with prematurity.^{21 22}

Visual acuity was generally good in both the FASD group and in the controls, but the range was wider in the FASD group. All participants managed to perform a visual acuity test correctly, but OCT measurements and visual field tests were sometimes hard to obtain in the FASD group. Several individuals with FAS could not perform a visual field examination due to attention deficits or difficulties with comprehending instructions. We think low IQ (median 70, range 36–121) and high prevalence of attention deficit hyperactivity disorder 21/26 (81%) in the FASD group likely contribute to this observation. Some participants with FASD also found it harder to perform the disc measurements than the macular measurements since they slightly had to move their gaze to find the fixation mark when performing the disc measurements. An adverse event that we had not expected was the discomfort from the OCT machine flashlight that some participants with FASD experienced.

Animal studies have shown that the formation of the eye is vulnerable to alcohol exposure.^{2 4} The sonic hedgehog pathway of molecular signals is central in the early development of the eye. If this pathway is disturbed, for example, through alcohol exposure in utero, it might cause microphthalmia.⁴ The third week of gestation is a critical period for alcohol-induced dysmorphology,²³ and as the eye development starts during the same period, it is likely to be affected as well. Animal studies on eye development under alcohol exposure have indicated that high doses of ethanol in early pregnancy in mice might reduce the number of optic nerve axons.²⁴ Eason and colleagues have found that zebra fish embryos exposed to alcohol have an increased apoptosis in the retina.²

In ONH, the number of axons are reduced whereas the supportive tissue of the nerve is normal.²⁵ Its cause is thought to be multifactorial, and it has been associated with risk factors such as smoking, alcohol use and/or prescription drug use.¹⁰ However, large epidemiological registry studies on humans are yet not available.²⁶ The prevalence of ONH differs between populations, ranging from 1.8 per 100 000 to 17.3 per 100 000.¹⁰ In a population of Swedish children, a prevalence of about 17 per 100 000 was reported.²⁷ The authors also found developmental problems in the cohort with ONH, but did not discuss any possible comorbidity with FASD. Another Swedish study found an increased risk of ONH if the mother was smoking during pregnancy, but could not find any data on maternal alcoholism in the 100 medical files included in the study. However, alcohol use in early pregnancy might not be captured in regular medical files.²⁸ Since alcohol consumption is associated with smoking, elevated risk of ONH associated with maternal smoking may be confounded by alcohol exposure, and conversely, risk associated with alcohol exposure may be confounded by maternal smoking.²⁹ In addition, whole genome sequencing has shown that ONH sometimes has an underlying genetic cause, and so causal attribution to risk factors must be done carefully.³⁰

In the present study, the ODA did not differ between the groups. It is important to note methodological differences between measuring ODA with an OCT machine and on a fundus photograph. Manual estimations of the optic disc via funduscopy or fundus photography assess a two-dimensional image, whereas an OCT machine through echo technique constructs a 3D image, also assessing structures below the disc surface when determining the ODA. Because ONH is a clinical diagnosis, it may not translate perfectly to a small ODA. Maybe, the ONH diagnosis should not only be detected clinically but confirmed with OCT showing subnormal peripapillary RNFL and/or in combination with a small ODA, reduced rim area and reduced BCVA. Figure 3 highlights this problem.

A previous study reported thinning of the RNFL as measured with OCT in adults with chronic alcohol and tobacco use, maybe as a result of optic nerve atrophy, but neither ONH nor optic nerve atrophy was assessed in the study.³¹ In our FASD group (n=26), 5 were smokers, 16 did not smoke and in 5 cases there was no information. One young adult suffered from alcohol addiction, 14 endorsed occasional binge drinking and 12 never drank alcohol. Only five neither smoked nor drank alcohol. Although these patterns largely reflect habits of young adults in the general population, possible additional negative effect of alcohol use and smoking on the RNFL in our cohort must be considered.

One strength of this study is that all FASD participants were recruited from a population-based sample of adoptees and diagnosed with the same diagnostic criteria.¹⁴ There are also some limitations. Researchers were not blinded to the status of the participants. The controls were predominantly medical students and not randomly selected and thus do not represent the general population. Further, smoking habits and alcohol use were not investigated in the controls.

In conclusion, significant differences in peripapillary and macular RNFL, retinal thickness, and BCVA were found in this group of young adults with FASD compared with healthy controls, and the results indicate that the RNFL thickness is decreased in individuals exposed to alcohol in utero. However, no difference in the size of the optic disc was found. As the RNFLs were found to be significantly thinner in the FASD group, an OCT examination is highly recommended when investigating individuals with FASD or suspected FASD in order to detect damage to the RNFL, as this may produce visual field defects and impact visual function. It remains to be investigated how structural differences of the visual system in individuals with FASD may affect visual perception as well as health-related and visual-related quality of life.