The quality and dimensions of the overlying tear film before, during, and after a blink influence the overall optical properties of the eye described by the higher order aberrations.1–6 Perturbations in the tear film leading to alterations in the propagation of light are expected to impact on the overall refractive properties of the tear film and, in turn, affect the overall quality of the retinal image. The effect of these perturbations is enhanced by the non-uniform distribution of lipid and aqueous layers, and by the difference in refractive index between them; lipid layer: 1.482, with a range of 1.460–1.5307 at a typical corneal surface temperature, and aqueous layer: 1.337.8,9

Abnormal tear film quality and structure due to dry eye disease leads to a reduced visual performance.10–12 Tear enhancers are intended to provide both relief from symptoms of dry eye and improve the stability of the tear film with a better quality of vision. These effects can also be produced by expressing meibomian oil from the lower eyelid by digital manipulation of the tarsal glands.13,14 Many patients report fluctuating and blurred vision immediately after tear enhancer instillation, or after meibomian oil expression. The visual disturbance can persist for many seconds before the quality of vision settles. Though such remarks are anecdotal, the patients are describing a persistent inconvenient symptom.

Considering the interruption in the optical quality of the eye, it would be useful to ascertain if any particular aberrations are more affected by deliberate shifts in the content and quality of the tear film. Using higher order aberration (HOA) analysis, it is possible to investigate whether specific numerical indices associated with the optical quality of the eye are more sensitive to changes related to the tear film.

The aim of this study was to determine if certain higher order aberrations observed from the corneal surface are related to: irregularities within and over the tear film, and whole eye; the quality of the retinal image quality; and are more susceptible to change shortly after instilling a tear enhancer or meibomian expression in normal subjects.

Assessment for data normality found that most of the corneal surface and whole eye HOAs were not distributed normally, and therefore non-parametric tests where applied to the data. The main significant results are listed in Tables 1–5 and Figs. 1–4.

Figure 1.

Selected data comparing RMS HOA data from right and left eyes. The two sets of data with the highest correlations are shown. All values are in microns (μ). The equations for the least squares regression lines are: corneal surface 3mm pupil (upper line), y=0.342x+0.052 (r=0.334, n=50, p=<0.05) and total ocular 3mm pupil (lower line), y=0.309x+0.020 (r=0.335, n=50, p=<0.05).

(0.05MB).

Although tear enhancers and meibomian expression are used in the treatment of dry eye disorders, the aim of this study was to identify the coefficients of higher order aberrations (HOAs) of the corneal surface and the whole eye that were most susceptible to change from the addition of these substances to the tear film of normal subjects.

While tear enhancer use in normal eyes has been previously investigated,20–24 there are only anecdotal comment reports relating to the effect of meibomian gland expression on the alteration of vision quality. The higher order aberrations, observed from both the corneal surface and whole eye, are useful in understanding the subjective effects of these treatment methods on vision by assessing the objective effects on optical quality.

The study found that instillation of a tear enhancer or the expression of meibomian oil had no effect on the mean and median value of each HOA for the smaller 3mm pupil. For the larger 6mm pupil, the tear enhancer reduced the magnitude of the total ocular aberrations (Z3−1, Z42, Z40, Z51, Z53 and Z55), indicating an overall improvement in optical performance across the subjects over the duration of the investigation. However, for the 6mm pupil, meibomian gland expression produced a significant increase in mean HOA values for the corneal surface aberrations (Z3−3, Z44, Z5−3, Z5−1), indicating that meibomian gland expression changes the regular distribution of tear film. In summary, manipulating the content of the tear film had a detectable impact on the short term optical performance at the larger 6mm pupil, but not for the smaller 3mm pupil.

These effects are evident in Fig. 2 and Tables 2 and 3, which show that the HOAs significantly changed after introduction of the tear enhancer and meibomian gland expression. It appears that, for the total whole eye, HOAs reduced after instillation of the tear enhancer, and, for the corneal surface, HOAs increased after meibomian gland expression.

Figure 2.

(a) Bar chart showing the RMS HOA where significant changes were detected after instillation of the tear enhancer. The appropriate standard deviations are shown as ‘T’ lines. All values are in microns (μ). Significant changes were uncovered for total ocular aberrations for the 6mm pupil only. There were no significant changes in any of the corneal surface aberrations. (b) Bar chart showing the RMS HOA where significant changes were detected following meibomian expression. The appropriate standard deviations are shown as ‘T’ lines. All values are in microns (μ). Significant changes were uncovered for the corneal surface aberrations Z3−3, Z44, Z5−3, Z5−1 and total ocular Z5−5 HOA for the 6mm pupil only. There were no significant changes in any of the total ocular aberrations.

(0.09MB).

From a clinical treatment perspective, tear enhancers are expected to raise the stability of the pre-corneal tear film, thereby reducing the rate of tear evaporation, without a negative impact on tear film optics. The fall in magnitude of the six specific HOAs after instillation of the tear enhancer observed in this study shows that this treatment method fulfils the third requirement. However, the expression of meibomian oil only fulfils the first two requirements,13,14 but not the third. This suggests that the expressed meibomian oil does not interact with the pre-corneal tear film in the 30s after the forced expression in a manner that promotes good optical performance. Local perturbations in the tear film caused by inadequate distribution of the expressed meibomian oils are the likely cause of an increase in mean values of particular corneal surface HOAs.

Table 4 lists the HOAs where the change in the value of the HOA was significantly associated with the initial RMS value for the HOA before instillation of the tear enhancer. Fig. 3 shows the data for the HOA with the highest correlation coefficient (Z5−1, corneal surface, 6mm pupil). The nature of the least squares line in Fig. 3 suggests that instillation of the tear enhancer will either reduce the HOA when the initial RMS value is high, or increase when the initial RMS value is low. Other HOA coefficients noted in Table 4 have similar effects to Z5−1, such as Z4−2 and Z5−3 (corneal surface, 3mm pupil). This implies that in some cases where the optical performance is relatively poor, the instillation of the tear enhancer is likely to improve overall optical performance. Conversely, in those cases where the optical performance is relatively good, the instillation of the tear enhancer is likely to reduce overall optical performance. Similar tendencies were found for other HOA correlations, such as Z51 (corneal surface, 3mm pupil) and Z5−1 (corneal surface, 6mm pupil). This has a practical application when measuring corneal topography in cases where the corneal surface HOAs are too high. In these cases, the tear enhancing drop is expected to improve the overall quality of the air-tear interface by reducing the HOAs, thus resulting in a more useful surface topographical analysis. This supports anecdotal reports of better quality corneal topography maps obtained in dry eyes or on irregular corneas after instilling tear enhancing drops.

Figure 3.

The HOA, Z5−1, where the highest correlation was found when comparing the change in the RMS value (Δ) with initial RMS value before instillation of the tear enhancer. The equation for the least squares regression lines is, y=0.107−0.942x (r=−0.871, n=50, p=<0.01. Excluding outlier, r=−0.774).

(0.05MB).

Table 5 lists the HOAs where the change in the value of the HOA was significantly associated with the initial RMS value for the HOA before meibomian gland expression. Fig. 4 shows the data for the HOA with the highest correlation coefficient, namely Z3−3 (total ocular aberration, 3mm pupil). A negative correlation was encountered in 19 of the 35 cases, where changes in the RMS value of specific HOAs were significantly associated with the RMS value before expression. In those cases where the initial value of the HOA was high, meibomian gland expression tended to reduce the HOA value, implying an improvement in optical performance. Conversely, in those cases where the initial value of the HOA was low, meibomian gland expression tended to increase the HOA value, implying a reduction in optical performance. In summary, poor optical performers tended to gain, good optical performers tended to lose out in the short term after meibomian gland expression.

Figure 4.

The HOA, Z3−3, where the highest correlation was found when comparing the change in the RMS value (Δ) with initial RMS value before meibomian expression. The equation for the least squares regression lines is, y=0.077−1.080x (r=−0.760, n=50, p=<0.01. Excluding outlier, r=−0.523).

(0.06MB).

No single coefficient stood out as the leading marker of change in the optical quality of the tear film after intervention. The RMS values of Z55 and Z40 indices describing the total ocular aberrations for a 6mm pupil are more susceptible to significant change when an improvement in optical performance after tear film intervention is envisaged. The Z3−3 and Z44 indices describing the corneal surface aberrations for a 6mm pupil are more susceptible to change when depreciation in optical performance after tear film intervention is anticipated.

Previous research has shown that the optical properties of right and left eyes are similar when these properties are quantified by the higher order aberrations.25,26 This is because the optical characteristics exhibited by the two eyes have a tendency towards symmetry.26 However, the inter-ocular differences in corneal surface curvature and refractive error suggest that there may yet be inter-ocular differences in the HOAs associated with the corneal surface and, possibly, the whole eye. A side-benefit of this study was the opportunity to further investigate these possible differences.

Our study found no measurable significant difference between the right and left eyes, implying that the optical profiles obtained from right and left eyes are the same in terms of basic statistics and drawn from the same population. This lack of any discernible inter-ocular difference supports previous findings.25,26 Using a binocular Shack-Hartmann aberrometer, Hampson and Mallen27 found inter-ocular correlations for coma, but for no other HOA. Our study also uncovered a similar trend in total ocular coma (Z31) for the 3mm pupil size, but not for the 6mm. In addition, we also encountered significant inter-ocular correlations for Z31 (corneal surface, 3mm pupil), Z3−3 (corneal surface, 6mm pupil), Z3−1 (corneal surface, 6mm pupil), and Z5−3 (corneal surface, 6mm pupil). Coma for the smaller 3mm pupil size, represented by Z31, was the corneal surface and total ocular HOA with the highest inter-ocular correlations, as shown in Fig. 1.

There are limitations in our investigation. We did not include any cycloplegia to control the effects of any micro-fluctuations of accommodation. This could limit the value placed on the effects on the whole eye HOAs results. However, micro-fluctuations of accommodation are more associated with their effects on lower order aberrations and have a lesser effect on HOAs.28,29 The subjects enrolled in this study were normal and not expected to demonstrate overtly poor corneal surface or whole eye performance. It could be that meibomian oil tended to reduce corneal surface optical performance because the subjects were normal. In those cases with relatively high values for specific HOAs, the HOA value tended to reduce after meibomian gland expression. Ridder et al.30 reported contrast sensitivity at high frequency (14cpd) to reduce after instillation of particular tear enhancers in normal eyes. Contrasting with this, a study on 5 cases of mild to moderate dry eye revealed an improvement in contrast sensitivity with a particular tear enhancer.20 The difference between studies on relatively dry and normal eyes may be associated with several factors, such as the compatibility and miscibility of the constituents of the tear enhancement regimen, or the tears themselves, and the unique optical properties of the individual eyes under investigation.

It has also been suggested that the dynamic behaviour of the tear film may impact on any correlation between the two eyes.31 Attempts were made to reduce the effects of tear film dynamics by restricting data acquisition to 5s after the final blink on all occasions. Nevertheless, even with this restriction, significant inter-ocular correlations were found in just 5 of the 60 sets of comparisons. Others have found that the optical performance of the eye tends to peak at approximately 6–7s after the blink before deteriorating.30,32 The difference in optical performance between 5 and 6s after the blink were almost imperceptible on closer examination of the reported data. Thus, harvesting data at 5s after the blink was not expected to impact on any inter-ocular correlations.

Lastly, the mean HOA values do not change significantly after adjusting the content of the tear film when the pupils are small. Significant shifts in the mean values can occur when the pupils are large. In practical terms, a patient with tear-related problems may report visual problems when the pupil is large, e.g. driving at night or near work under poor light. In the standard clinical setting, visual problems may not be picked up using standard low contrast charts, because the ambient lighting limits the effect the pupil has reducing the HOA values. However, objective assessment of HOA values under larger pupil conditions may lead to a better understanding and association with reported symptoms.

In conclusion, this study found no detectable differences between right and left eyes when using a Shack–Hartmann aberrometer. When comparing the HOAs values obtained before intervention and 30s afterwards, no change in optical quality was noted for the 3mm pupil, but for the 6mm pupil the tear enhancer tended to improve optical performance of the whole eye, and meibomian gland expression tended to reduce the optical performance at the corneal surface. Further studies are needed to investigate the effect of these two treatment methods on subjects with abnormal tear film characteristics due to dry eye.

The authors have no conflicts of interest to declare.