G Model
ARTICLE IN PRESS
CLAE-772; No. of Pages 8
Contact Lens & Anterior Eye xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Contact Lens & Anterior Eye journal homepage: www.elsevier.com/locate/clae
A new model of central lid margin apposition and tear film mixing in spontaneous blinking Heiko Pult a,b,∗ , Donald R. Korb c , Paul J. Murphy d , Britta H. Riede-Pult a,b , Caroline Blackie c a
Dr. Heiko Pult – Optometry & Vision Research, Germany Cardiff University, School of Optometry & Vision Sciences, UK c Korb Associates and TearScience Inc., USA d University of Waterloo, School of Optometry & Vision Science, Canada b
a r t i c l e
i n f o
Article history: Received 13 August 2014 Received in revised form 21 January 2015 Accepted 26 January 2015 Keywords: Over-blink Spontaneous blink Tear film Tear film spreading Central lid position Spontaneous blinks Dry eye
a b s t r a c t Purpose: To investigate tear film spreading and central lid position in spontaneous blinks and to propose a model of the central lid position of the closed eye in such blinks. Method: In vivo: lid margin position and geometry of 15 subjects (9 female; median age = 45 years) were evaluated by high-speed video and slit-lamp microscope video in consecutive spontaneous blinks. Upper lid (UL) tear meniscus (TM) depth was observed in the open and almost closed eye. Eyelid geometry, position and UL TM depth were analysed by Image-J Software. Lid margin thicknesses were measured with a Scheimpflug camera. In vitro: tear film spreading and lipid layer formation were simulated on a lubricated glass plate and videoed by high-speed camera (JVC, GZ-GX1BE, Japan). Results: In vivo: the median central lid margin thickness was not significantly (p = 0.258) different between UL (1.8 mm) and LL (1.7 mm) in the opened eye. During blinking, UL remained perpendicular to the corneal surface, while LL tilted in and thinned. A scaled model diagram was created and revealed an over-blink of the UL over the LL (>0.7 mm) and a height offset of the posterior lid margin of >0.7 mm. In vitro: the LL TM fused with the UL TM even before full lid touch due to capillary bridge building. Conclusions: The central UL overlaps the central LL during spontaneous blinking. This provides the appearance of complete closure. The space that results from the lack of lid margin apposition influences the fusion of the upper and lower TM and ultimately tear film mixing. © 2015 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved.
When a human eye opens fully, in primary gaze, the upper eyelid margin is moved upwards and the surface of the margin is angled forwards, while the lower lid margin is moved downwards and the margin surface angled forwards [1]. During eyelid closure, the orbicularis muscle pulls the upper lid down towards the lower lid and the lower lid is pulled slightly nasally [2,3]. Bron et al. [4] state that during lid closure the occlusal surfaces of the lid margins are apposed. There appears to be an inherent assumption in our understanding of the blink that during lid closure, the lid margins, specifically the occlusal surfaces, come into contact [5,6]. This apposition results in fusion of the tear menisci and lipid reservoirs along both lid margins [7,8]. In the inversion of this argument, any fusion of the tear menisci and lipid reservoirs along the lid margins
∗ Corresponding author at: Dr. Heiko Pult – Optometry & Vision Research, Germany. Tel.: +49 1749025090; fax: +49 6201873657. E-mail address: [email protected] (H. Pult).
would not be possible without sufficient lid touch. Based on this assumption, it was hypothesised that there must be some rotation of the upper lid margin surfaces during blinking to bring the lid margins into apposition [1]. However, recent evidence has demonstrated that the central portion of the upper and lower lid margins frequently do not touch during spontaneous blinking [5,9]. Further, an analysis of highspeed video recording of the blinks demonstrated an over-lapping of the central upper lid over the central lower lid, named as an ‘overblink’. This over-blink likely results in a space posterior to the inner upper lid as the central upper lid margin rests over the lower lid lashes. It has been suggested that the over-blink is arguably the normal state in healthy Caucasians and a possible reason for the observation of an unaltered drop of vital dye placed on the lower lid margin during blinking [9]. These observations raise two questions: (i) what positions do the lid margins actually take in a completed, spontaneous blink; and
http://dx.doi.org/10.1016/j.clae.2015.01.012 1367-0484/© 2015 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
G Model CLAE-772; No. of Pages 8
ARTICLE IN PRESS H. Pult et al. / Contact Lens & Anterior Eye xxx (2015) xxx–xxx
2
(ii) how does tear film mixing occur, leading to the establishment of a proper lipid layer. Since the over-blink is likely the normal state in healthy Caucasians, the need for full lid touch during a blink for proper tear film mixing is called into question. We hypothesise that even within this small space between central the lids, both tear menisci can fuse during blink lid closure due to rheological mechanisms. However, to better understand these mechanisms, more detailed information about the central lid position of the closed eye in spontaneous blinks is needed. Furthermore, possible rheological mechanisms behind this new blink model need to be investigated. Having answered these questions may give us a better understanding of tear film mixing and lipid layer formation in the over-blink lid position, since the anecdotal concept of tear menisci fusion based on the lid touch may not be relevant. The purpose of this pilot-study was to investigate these questions by examining tear film spreading and central lid position of the closed eye in spontaneous blinks in a cohort of human subjects, and to propose an in vitro model of the central lid position of the closed eye in complete, spontaneous blinks.
Fig. 1. Sketch showing the different camera positions (1: temporal view; 2: inferiortemporal view; 3: frontal view).
1. Method
1.2. Techniques (Table 1)
For the human study, 15 healthy subjects (9 females; median age = 45 years) were recruited from the clinical practice Horst Riede GmbH, Weinheim, Germany. As has been previously described elsewhere, a 0.15 l lissamine green drop was placed on the anterior portion of the central lower lid margin of the lower lid (LL) [5,9]. Each subject was asked to blink normally for 30 s, and lid margin position and geometry were evaluated by high-speed video and slit-lamp microscope video. UL tear meniscus depth (z-axis) was observed in the completely opened eye and almost closed eye using slit-lamp photography. The term ‘almost closed’ describes the smallest possible lid aperture in which the lid margins still were observable. In this position the UL and LL margins still show a vertical gap of up to half of the UL margin’s thickness (from the temporal-inferior view) [10]. Eyelid geometry, margin position and upper lid tear meniscus depth were analysed by Image-J Software (Version 1.42q; Wayne Rasband, National Institute of Health, Bethesda, MD, USA) from images taken from the video. Lid margin thicknesses and corneal topography were measured with a Scheimpflug camera (Pentacam, Oculus Optikgeräte GmbH, Wetzlar, Germany). Based on these data, a scaled model diagram was created to represent the central lid position of the closed eye in spontaneous blinking. All observations were done by the same observer (HP).
1.2.1. Lissamine green drop test This test was used to exclude touch of the central lids in spontaneous, complete blinks and to evaluate the minimum distance between the UL and LL margins in blinks. Subjects were positioned on the chin-rest of the slit-lamp microscope and a 0.15 l lissamine green drop was placed onto the anterior portion of the central, LL using a microliter pipette (neoLab-Mikroliterpipette Economy, neoLab Migge Laborbedarf-Vertriebs GmbH, Heidelberg, Germany). The anterior portion was defined as extending from the meibomian gland orifices to the eyelash line [5]. The lissamine green drop was photographed at each subject’s lid margin and the drop height was measured using Image-J software. 1.2.2. Video Subjects were asked to maintain normal spontaneous blinking over a 30 s period while looking straight ahead. A frontal view of blinking was digital recorded using a video slit-lamp microscope (DigiPro2, bon Optic, Lübeck, Germany). Simultaneously, the same blinks were videoed from a temporal-inferior view (Fig. 1) using a high-speed video camera (250 frames/s; JVC, GZ-GX1BE, Japan). A previous investigation revealed that the temporal-inferior view provided the best observation of the over-blink [9]. The extent of over-blink was analysed subjectively on a 5-grade scale. Additionally, blinks were videoed at high-speed from a side-view to analyse lower lid movement. Number of completed, spontaneous blinks were recorded; any potential non-spontaneous blink was excluded by observing Bell’s phenomenon.
1.1. Exclusion criteria Subjects were excluded if they had floppy eyelids [11] (easily everted, floppy upper eyelid), conjunctivo chalasis, irregular lid margins (such as irregularity in, for example, cicatricial meibomian gland dysfunction), supra-nuclear motor impairment of lid movements, ptosis, apraxia, rheumatoid arthritis, diabetes, recent ocular infections, hay fever, any history of ocular surgery, use of any medication or eye drops known to affect the ocular surface, worn contact lenses or had contact lens experiences, or were pregnant. All procedures were conducted in accordance with the Declaration of Helsinki (2000), and approval for the study was given by the Cardiff School of Optometry and Vision Sciences Ethics Committee. All subjects were fully consented prior to study enrolment.
1.2.3. Geometry and position of the central eye lids Upper eyelid: the relative position of the UL to the cornea was measured from the high-speed video recordings using Image-J software. Imaging of the lid margins of the opened eye were captured by Scheimpflug camera (Pentacam, Oculus, Wetzlar, Germany) and central lid margin thickness (distance between eye lash line and cornea) and angle of the lid margin surface, relative to the cornea, were measured from the images using the analysis tool of the Pentacam software. Variations of the UL margin thickness during each blink were observed by measuring the distance between the eyelash line and the UL tear meniscus, as well as the distance between the eyelash line and cornea. The position of the UL were photographed in the fully opened eye and almost closed eye using the slit lamp microscope, assisted by the use of a small dental mirror
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
G Model CLAE-772; No. of Pages 8
ARTICLE IN PRESS H. Pult et al. / Contact Lens & Anterior Eye xxx (2015) xxx–xxx
3
Table 1 Summarising table of the different techniques. Technique
Purpose
Procedure
(i) To exclude touch of the lid margins in complete, spontaneous blinks (ii) To analyse the minimal distance between central lid margins after completed, spontaneous blink
(i) 0.15 l lissamine green drop placed onto the anterior portion of the central, LL (ii) Measurement of the lissamine green drop height
To analyse central lower lid vertical movement, eye ball retraction and bells phenomenon
High-speed video from the temporal view (camera position 1) and digital image analyses
(i) To classify the over-blink (ii) To analyse the relative position of the UL to the cornea in complete, spontaneous blinks
High-speed video from the inferior-temporal view (camera position 2) and subjective and digital image analyses
Frontal view photography
To analyse lissamine green drop height
Digital slit lamp microscope observation (camera position 3) and digital image analyses
Inferior view photography
To measure variations of the UL in the opened and almost closed position
Digital slit lamp microscope plus dental mirror to observe the UL margin and digital image analyses
Scheimpflug camera
To analyse central lid margin thickness and relative eye lid position to the cornea of the opened eye
Scheimpflug camera observation and digital image analyses
Glass plate model
To simulate tear film spreading and tear meniscus fusion
Glass plate model with saline solution and different coloured dyes. High-speed videoing and digital image analyses
Lissamine green drop test
Video: side view
Video: inferior-temporal view
positioned to permit viewing of the eyelid margin from below for the almost closed eye. The mirror helped to avoid any perspective error that would make the almost closed UL appear thicker when viewed from the front. Lower eyelid: to measure the LL position, a small horizontal slit beam from the slit lamp illumination was projected onto the central LL margin, directly below the centre of the subject’s pupil. Since full closure of the UL would hinder the view of the LL margin, the subject’s eyebrow was gently lifted by the patient during measurements. Although only a slight movement on the eyelid was produced, this was sufficient to assist in viewing of the LL, while not interfering with normal lid movement. Blinks were again highspeed videoed from a side view, and central LL vertical movement were measured using the position of the projected small horizontal light slit with Image-J software. Finally, any retraction of the eyeball during blinks was analysed by Image-J software by observing the relative movement of the eyeball.
1.2.4. Over-blink The over-blink of the UL during complete blinks was evaluated by analysing the high-speed video (inferior-temporal view) and classified subjectively (Fig. 2) using a 5-grade scale [9] (0: aligned; 1: over-blink of UL extending less than 1/3 of UL margin thickness, 2: over-blink of UL extending between 1/3 and 2/3 of UL margin thickness, 3: over-blink of UL extending between 2/3 and 3/3 of UL margin thickness, 4: over-blink of UL extending more than 3/3 of UL margin thickness).
Fig. 2. Grading scale of the over-blink [9].
1.2.5. Upper lid tear meniscus At least 30 min after observations described above, the depth (zaxis) of the UL tear meniscus was photographed in the fully opened eye lid position and almost closed eye lid position using the slit lamp microscope and a small dental mirror. The tear meniscus was coloured by fluorescein to assist in observation. To allow repeatable fluorescein application, the fluorescein strip (Bio Glo, Biovision Ltd., Dunstable, UK) was modified according to Pult et al. [12]. In summary, a normal fluorescein strip was modified by folding the dye impregnated tip at the 1 mm length (named the modified fluorescein strip (MFS)), which was then moistened with a single drop of saline (Software Saline, Alcon, Großostheim, Germany) and shaken to remove excess fluorescein before instillation. 1.2.6. Glass plate model For the in vitro model study, tear film spreading and lipid layer formation were simulated on a glass plate model lubricated with saline solution (Software Saline, Alcon, Großostheim, Germany; concentration: 0.9% (based on manufacturer’s data)) (Figs. 3 and 4). Two glass plates, representing the lids, with a thickness of 1.9 mm each, were placed on a glass plate base. The thickness of the glass plates was measured using a calliper rule. The base plate was loaded with the saline solution and a small lissamine green drop was placed at the tear meniscus of the glass plate representing the lower lid. Blinks without ‘lid-touch’ were simulated by moving the second glass plate (upper lid) to and from this first glass plate (lower lid), without touching, and videoed by high-speed camera. The model was then observed to see, if after at least one simulated “noncontacting” blink, the lissamine green had spread across from the upper lid to the lower lid glass plates. To then simulate mixing of the lipid layer, red coloured oil (St. John’s Wort red oil, Dr. L. Schmitt GmbH & Co. KG, Planegg, Germany) was placed on top of the saline layer, along with a small green oil drop (pumpkin seed oil, Fausere Vitaquellwerk KG, Hamburg, Germany) placed at the tear meniscus of the lower eyelid glass plate. Potential mixing of the tear film was then evaluated subjectively by simulating blinks using movement of the glass plates. In addition to subjective evaluation, the glass plate model was videoed from a side view and loaded with the minimum volume of saline solution to achieve full wetting of the glass plate base
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
G Model CLAE-772; No. of Pages 8
ARTICLE IN PRESS H. Pult et al. / Contact Lens & Anterior Eye xxx (2015) xxx–xxx
4
Fig. 3. Simulation of aqueous tear film mixing in the glass model (lubricant: saline solution; one tear meniscus was coloured by lissamine green). Left: simulated closing phase of a blink. Middle: completed closing phase without ‘lid-touch’ (=return point). Right: opening phase of the simulated blink. Lissamine green was taken from the lower lid glass plate and is distributed over the lid aperture and is already in the upper lid glass plate tear meniscus.
Fig. 4. Simulation of lipid layer mixing in the glass model. (A) Model is lubricated by saline solution. Small green oil drop (black arrow) at the lower lid glass plate meniscus and red oil spread over the total ‘tear-film’ (dashed arrow). (B) Simulated closing phase of a blink. (C) Opening phase of the simulated blink. Green oil drop is slightly mixed with the red oil (dashed arrows).
and to produce visible tear menisci along the glass edges of both overlying plates. A subjective assessment was then made whether tear meniscus fusion would occur, depending on the minimum and maximum possible saline load of the model. The glass plate experiments were trained several times before the experiment, especially to guarantee repeatability and appropriate simulated blink speed. In the experiment the achieved simulated blink speed and the distance until first tear menisci fusion was analysed from the videos of 5 consecutive experiments. 1.2.7. Statistical analyses Data were examined for normality by Shapiro–Wilk test. Correlations between measurements and the over-blink (ordinal scale, non-parametric) were evaluated by Spearman rank. Differences between lid thicknesses (parametric) were evaluated by paired t-test. The data were analysed using SPSS 20.0 (SPSS Inc., Chicago, USA) and WinStat 2007.1 (R Fitch Software, Bad Krozingen, Germany). 2. Results 2.1. Human studies The lissamine green drop was observed to remain unaltered for a median of 8 consecutive, complete blinks for all subjects (Table 2). The observed median over-blink grade was 3 (interquartiles: 3–4). The eyeball was drawn back into the orbit during blinking by a median distance of 0.9 mm (interquartiles: 0.36–0.97). The median height of the instilled lissamine green drop at subject lid margins was 0.4 mm (y-axis; interquartiles: 0.37–0.47) and such slightly higher than expected of a 0.15 l drop. Since the drop was not altered after consecutive blinks, the vertical space between the
upper and lower central anterior lid margins must have been at least 0.4 mm in the closed eyelid position. The median central lid margin thickness was slightly, but not significantly (p = 0.258), larger for the upper lid (1.8 mm) than for the lower lid (1.7 mm) when the eyelids were open. The upper lid margin was also observed to remain perpendicular to the corneal surface at all times during blinking. The lower lid margin tilt changed from the fully open position by 23◦ (median, upwards, inwards) and thinned by 20.0% (Table 2). This change in thickness was significantly correlated to the over-blink score (r = 0.791; p = 0.0032). The distance between the upper lid eyelash line and cornea also increased by 1.17 times (median) from the fully opened position to the almost closed position. This increase was significantly correlated to overblink scores (r = 0.779; p = 0.002). Therefore the median distance between cornea and eyelash line of the upper lid appears to be larger in the closed eyelid position. Concurrently, the distance between the upper lid eyelash line and the outer margin of the upper lid tear meniscus only enlarged by 1.09 times (median), between the fully opened position and the almost closed position. This parameter was significantly correlated to over-blink scores (r = 0.553; p = 0.039). At the same time, the tear meniscus depth increased by a median ratio of 1.8, which was also significantly correlated to over-blink scores (r = 0.875; p < 0.001). In other words, in the almost closed eye, the upper eyelid appears to thicken and move away from the corneal surface (cease to touch the cornea), with the space created being filled by the tear film, revealed by an expanded tear meniscus depth. To further investigate these changes, a scale diagram of the lid positions was constructed from the evaluated data (Fig. 5). Observation revealed an over-blink of the UL over the LL by >0.7 mm (z-axis) and a height offset of the posterior lid margin of >0.7 mm
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
G Model CLAE-772; No. of Pages 8
ARTICLE IN PRESS H. Pult et al. / Contact Lens & Anterior Eye xxx (2015) xxx–xxx
5
Table 2 Median values and interquartiles of the observed parameters. Means and standard deviation (SD) are additionally shown in parametric data, only. Parameter
Median
In vivo experiment Number of consecutive spontaneous complete blinks Draw back of the eye ball [mm] Lissamine green drop height [mm] Over blink Upper lid thickness (opened) [mm] Upper lid thickness (closed) [mm] Lower lid thickness (opened) [mm] Lower lid thickness (closed) [mm]
8 0.9 0.4 3 1.8 2.1 1.7 1.4
In vitro experiment Distance between glass plates at tear menisci fusion [mm] Simulated glass plate speed [cm/s]
0.5 23
Mean
±SD
7–13 0.35–0.97 0.37–0.47 3–4 1.56–1.90 1.65–2.25 1.56–1.90 1.23–1.55
9 0.7 0.4 n/a 1.7 2.0 1.7 1.4
3.79 0.14 0.12 n/a 0.22 0.25 0.24 0.23
0.35–0.65 15.8–23.0
0.5 20
0.16 3.92
Interquartiles
Fig. 6. Simulation of the menisci fusion without lid touch (saline solution, red vertical mark on glass plates = reference mark: 5 mm increments, dashed lines show tear meniscus height). Top: tear meniscus formation at both glass plates. Right glass plate is moved towards the left one. Middle: capillary bridge formation even distance between tear menisci was still 0.23 mm. Below: very thick capillary bridge when distance between tear menisci was zero.
Fig. 5. Scale diagram of lid position in the fully opened eye and closed eyelid positions, during spontaneous, complete blinks (median values are shown).
(y-axis), which produced a triangular shaped space between the cornea and lid margins. 2.2. In vitro The LL tear film meniscus was observed to fuse with the UL tear meniscus (Figs. 3 and 4) even before full lid touch (median distance between the glass plates was 0.5 mm in a simulated complete blink). Similarly, lipid spread occurred without any ‘lid-touch’, i.e. no contamination of the lower lid glass plate with coloured fluid from the UL glass plate during the opening phase. Simulated blink
speed was 23 cm/s (median speed of 5 consecutive experiment setups (Table 2)), which is close to reported figures in humans (20 cm/s [13]). To observe tear film fusion from a side view, the UL glass plate was moved towards the LL glass plate (∼5 cm/s, Fig. 6). This demonstrated that both tear menisci quickly build up a capillary bridge (Fig. 6). Subjectively, this fusion of the tear menisci occurred earlier (at a larger distance between plates) with maximum saline loading compared to minimal loading of the glass plates. Furthermore, it appeared that capillary bridge building developed shortly before the tear menisci outer limits met. Thus, for the tear menisci to merge, the lids do not need to meet, but just approach by a certain distance (0.23 mm in the glass plate model, Fig. 4). For irregular glass plate margins, multiple capillary bridges occurred, which enabled partial tear film mixing (Fig. 7), depending on the tear film volume (Fig. 8). The resting time of the plates in close proximity to each other also seemed to affect total tear film mixing, presumably due to changes in surface tension across the capillary bridge, and subsequent capillary flow.
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
G Model CLAE-772; No. of Pages 8
ARTICLE IN PRESS H. Pult et al. / Contact Lens & Anterior Eye xxx (2015) xxx–xxx
6
Fig. 7. Left image: initial development of capillary bridges between the two glass plates at three locations (right glass plate lubricant: fluorescein in saline solution; left plate: saline solution) showing partial tear film mixing. Right image: Simulated opening phase of blink.
3. Discussion This study has confirmed the results of previous investigations in Caucasians [5,9] and found that a drop of vital dye placed on the central lower eyelid margin remained unchanged after several (median, 8) consecutive, apparently complete, spontaneous blinks. An over-blink was also observed, with a median subjective grade of 3. This was slightly smaller than the expected over-blink calculated from the scaled model diagram. It was also observed that, while the UL margin swept normal to the corneal surface during blinking, the margin also thickened as it approached the LL margin for apparent closure. As the LL rotated inwards and moved horizontally towards the nose, the margin itself also thinned slightly by 0.3 mm (median) or 20%. The lissamine green drop with an median height of 0.4 mm was unaltered after normal spontaneous blinks, it can be assumed that the central lid margins of the UL and LL do not make direct contact, and that the minimum distance between the UL and LL is larger than the lissamine drop height. These changes in lid thickness during blinking, along with the UL over-blink placement, resulted in a small space between the posterior, central, UL near the margin and the corneal surface of at least 0.7mm in this cohort of patients. Furthermore, the scale model diagram illustrates that the lids appear to be fully closed from a frontal view, as commonly observed in humans [14–17]. The LL thinning is likely the result of the inward, nasal movement of the eyelid during spontaneous blinking [18,19]. However, the apparent thickening of the UL margin during blinking appears to be new information. This phenomenon could be a relative weakness of the orbicularis oculi muscle, particularly of the pretarsal portion, resulting in a slight lifting off of the lid wiper away from the cornea [9], or, more likely, due to muscle thickening during muscle contraction of the pretarsal portion of the orbicularis oculi muscle pulling the eyelid away from the cornea. There may also be an effect from the modulus of elasticity of the lid wiper portion of the eyelid, which is compressed in the fully opened eye, but becomes more relaxed during a blink [20]. Although the study conditions necessary for observation meant that these changes were seen when the subject was asked to make a voluntary blink to achieve an almost closed eye, it seems reasonable to assume that this situation will also occur during spontaneous blinking. The increased distance between the eyelash line and cornea, and the increased tear meniscus, support the hypothesis that the posterior eyelid margin contour changes during a blink. While Kessing’s space [21], between the tarsal UL and bulbar conjunctiva, has long been recognized, little is known regarding the dynamics of this space during blinking. Specifically, changes to the lid wiper contour
and the interaction between the cornea and lid wiper during blinking have not been previously documented [20,22]. One possible hypothesis is that the modulus of the lid wiper is promoting the enlargement of the distance between cornea and anterior lid margin at the almost complete eye lid position compared to the opened one, since lid pressure may be different between both positions [20]. Also, although the central UL margin remains perpendicular to the cornea throughout the blink, its final position in the closed eye appears to be angled inwards by approximately 64◦ , relative to a normal to the centre of the cornea. Similarly, the central lower lid tilts inwards by approximately 23◦ , relative to the normal. To our knowledge this is the first time that the angular position of both central eyelid margins has been reported in spontaneous blinks. However, the measurements used to produce the scale model diagram are limited by the available techniques. To our knowledge there is no technique available that allows the observation of lid position in spontaneous blinks non-invasively and with a high-speed sectional imaging. The scaled diagram of this pilotstudy is also based on a small population of mid-aged Caucasians and it would be of interest whether there is a difference between gender, races, or between symptomatic and asymptomatic dry eye patients in a larger population. Nevertheless, observations are well correlated to subjective over-blink scores showing that the scaled diagram approximates what really happens in such observed blinks. Therefore the triangular space between the closed eyelids is of significant interest and deserves continuing investigation. The model may also illustrate some practical benefits from not having full, parallel touch of the lids – for example, if it did occur, the tear film may be squeezed out, the meibomian gland orifices may block each other, or there may not be sufficient space for the lipid layer to compress into with a full blink. Since the results of the in vivo experiment demonstrated that the final tear film characteristics are not dependent on complete central lid margin apposition, tear film mixing was evaluated by using glass plates to simulate blinking, lubricated by saline solution. This in vitro experiment demonstrated that the two tear menisci do not need to overlap to mix. This is possibly due to the formation of a capillary bridge. The side view of the glass plates demonstrated this capillary bridge between the UL glass plate tear meniscus and the LL glass plate tear meniscus. The energy of capillary forces in fluid rheology are described as being very high, mainly depending on the lubricant surface tension [23], making it a potentially important mechanism in tear film mixing, at least for the glass plate blink model. To create a capillary bridge, both tear menisci must be gradually moved closer together, until they reach a certain critical distance
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
G Model CLAE-772; No. of Pages 8
ARTICLE IN PRESS H. Pult et al. / Contact Lens & Anterior Eye xxx (2015) xxx–xxx
7
Fig. 8. Difference in capillary bridge building produced by varying saline solution loading of glass plates; lower load (left) in comparison to higher load (right). The white arrows show likely flux of the fluorescein from beneath the glass plates towards space between plates (lid aperture).
apart. In the glass plate model it appeared that the extending edges of each tear meniscus do not even need to touch (Fig. 6) for capillary bridge building to occur shortly before touch. Capillary bridge building was related to loading of the model, therefore the tear film volume and lid geometry may have a great impact in spontaneous blinks. After any blink, the capillary bridge will collapse in the opening phase once a certain distance has been created between the lid margins and tear flux [24,25] will draw the tear film into the tear film menisci. Since central lid margins do not commonly touch during a blink and tear film volume appears to impact tear meniscus mixing in the closed eyelid position in spontaneous blinks, tear film volume and eye lid position in such blinks may be vital for the formation of a capillary bridge [26]. However, even if central tear menisci are not brought close enough for full bridging to occur, para-central capillary bridges may still be able to promote sufficient tear film mixing (Figs. 7 and 8). The glass plate model of the in vitro experiment is an approximation of the real eye, especially in terms of tear film viscosity, ocular surface topography and eye lid tension. However, the in vivo observation of the para-central capillary bridge building (Fig. 9) may demonstrate that the rheological mechanisms observed by the in vitro experiment are transferable to the real eye. It has been previously noted that tear volume may play a role in promoting lipid layer formation [27], thus, it is possible that this bridging phenomenon assists in the uptake of lipids from the lipid reservoir. Assuming a normal tear meniscus height range (in the open eye) for the UL of 0.17 mm [28] to 0.27 mm [29] and LL of 0.20 mm [28] and 0.31 mm [29], and comparing it to the distance between the posterior lid margins, modelled as being at least 0.7 mm, the distance between the two tear menisci would seem to be too large to allow for capillary bridging. However, observation of the UL tear meniscus in the almost closed eye showed that the meniscus enlarges, probably due to the tear film thickening as the same volume covers a smaller ocular surface area. This will increase tear volume along the lid margins and cause the tear menisci to extend further away from the margins in the very first moment of the completed blink, thereby reducing the distance between the two menisci and permitting capillary bridging. A space between the lid margins and the cornea would facilitate this, just as what appears to occur with the over-blink. Further investigation of these effects
Fig. 9. In vivo simulation of a capillary bridge between temporal tear menisci coloured by fluorescein. Top: early stage of capillary bridge formation (white arrow). Middle: full para-central capillary bridge (dashed arrow) after further minimal reduction of lid aperture. Below: sketch of full capillary bridge. Upper tear meniscus height next to the capillary bridge was 0.18 mm, the lower lid tear meniscus height was 0.17 mm and the distance between both was 0.30 mm.
in dry eye patients and contact lens wearers would be of great interest. The capillary force due to a liquid bridge, so called capillary bridge, and the breaking of this bridge, are of high importance due to the widespread existence of capillary menisci in macro-, micro-, and even nano-scale applications [30]. The capillary bridge facilitates the fusion of both tear menisci even if the central UL and LL tear menisci are not overlapping, as is probably the case in the over-blink. Until now, tear film fusion was considered to be possible only with lid touch in blinks. This appeared to be especially important in lipid layer formation [7,8]. In contrast, the over-blink is assumed to be the normal state in healthy Caucasians [9]. Furthermore, the investigation of this paper had shown that there appears to be a noticeable space between the posterior lid margins, and
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
G Model CLAE-772; No. of Pages 8
ARTICLE IN PRESS H. Pult et al. / Contact Lens & Anterior Eye xxx (2015) xxx–xxx
8
consequently between the UL and LL tear menisci. All of this calls the importance of the lid touch in tear film mixing into question. The observation of the capillary bridge formation may be the missing piece of information in tear film fusion in spontaneous, complete blinks (Fig. 9). This model for tear mixing using capillary bridging in the space created by the over-blink might be altered if the capillary bridging was impeded in some way, perhaps by the presence of lid parallel conjunctival folds (LIPCOF). LIPCOF scores are increased in symptomatic dry eye patients, and they are correlated to lidwiper epitheliopathy [31–34]. LIPCOF may physically hinder the blink action as it approaches eyelid closure [10] and they also influence tear meniscus distribution along the lid margin [35]. Therefore in patients with an increased LIPCOF score, insufficient para-central capillary bridge formation is likely, and tear film mixing, spreading and lubrication of the bulbar conjunctiva may be insufficient. This effect may only be dominant in the opening phase of a blink, but this phase may be vital for lipid layer spreading. Such an effect remains to be confirmed. 4. Conclusions The results of this study show that, in contrast to prior assumptions regarding lid apposition during blinking, the central upper lid margin overlaps the central lower lid margin during spontaneous blinking. This overlap of the upper lid margin provides the appearance of complete closure. The space that results from the lack of lid margin apposition influences the fusion of the upper and lower menisci and, ultimately, tear film mixing. Financial support None. Conflict of interest No conflicting relationship exists for any author. References [1] Foulks GN, Bron AJ. Meibomian gland dysfunction: a clinical scheme for description, diagnosis, classification, and grading. Ocul Surf 2003;1:107–26. [2] Bell C, Bell C. On the motion of the eye, an illustration of the uses of the muscles and nerves of the orbit. Philos Trans R Soc Lond 1823;113:166–86. [3] Doane MG. Blinking and the mechanics of the lacrimal drainage system. Ophthalmology 1981;88:844–51. [4] Bron AJ, Yokoi N, Gaffney EA, Tiffany JM. A solute gradient in the tear meniscus. I. A hypothesis to explain Marx’s line. Ocul Surf 2011;9:70–91. [5] Korb DR, Blackie CA, McNally E. Evidence that the upper and lower lids do not make complete contact even when the lids are closed. Cornea 2013;32:491–5. [6] Bron AJ, Yokoi N, Gaffney EA, Tiffany JM. A solute gradient in the tear meniscus. II. Implications for lid margin disease, including meibomian gland dysfunction. Ocul Surf 2011;9:92–7.
[7] Bron AJ, Tiffany JM. The meibomian glands and tear film lipids. Structure, function, and control. Adv Exp Med Biol 1998;438:281–95. [8] Bron AJ, Tiffany JM, Gouveia SM, Yokoi N, Voon LW. Functional aspects of the tear film lipid layer. Exp Eye Res 2004;78:347–60. [9] Pult H, Riede-Pult B, Murphy PJ. A new perspective on spontaneous blinks. Ophthalmology 2013;120:1086–91. [10] Pult H, Riede-Pult BH, Murphy PJ. The relation between blinking and conjunctival folds and dry eye symptoms. Optom Vis Sci 2013;90:1034–9. [11] Culbertson WW, Ostler HB. The floppy eyelid syndrome. Am J Ophthalmol 1981;92:568–75. [12] Pult H, Riede-Pult BH. A new modified fluorescein strip: its repeatability and usefulness in tear film break-up time analysis. Contact Lens Anterior Eye 2012;35:35–8. [13] Doane MG. Interactions of eyelids and tears in corneal wetting and the dynamics of the normal human eyeblink. Am J Ophthalmol 1980;89:507–16. [14] Doughty MJ. Further assessment of gender- and blink pattern-related differences in the spontaneous eyeblink activity in primary gaze in young adult humans. Optom Vis Sci 2002;79:439–47. [15] Doughty MJ. Consideration of three types of spontaneous eyeblink activity in normal humans: during reading and video display terminal use, in primary gaze, and while in conversation. Optom Vis Sci 2001;78:712–25. [16] Bentivoglio AR, Bressman SB, Cassetta E, Carretta D, Tonali P, Albanese A. Analysis of blink rate patterns in normal subjects. Mov Disord 1997;12:1028–34. [17] Collins M, Heron H, Larsen R, Lindner R. Blinking patterns in soft contact lens wearers can be altered with training. Am J Optom Physiol Opt 1987;64: 100–3. [18] Casse G, Sauvage JP, Adenis JP, Robert PY. Videonystagmography to assess blinking. Graefe’s Arch Clin Exp Ophthalmol 2007;245:1789–96. [19] Spooner J. Ocular anatomy. London: The Hatton Press Ltd.; 1957. [20] Jones M, Fulford G, Please C, McElwain D, Collins M. Elastohydrodynamics of the eyelid wiper. Bull Math Biol 2008;70:323–43. [21] Kessing SV. A new division of the conjunctiva on the basis of X-ray examination. Acta Ophthalmol (Copenh) 1967;45:680–3. [22] Dunn AC, Uruena JM, Huo Y, Perry SS, Angelini TE, Sawyer WG. Lubricity of surface hydrogel layers. Tribol Lett 2013;49:371–8. [23] Persson BNJ. Sliding friction. 2nd ed. Berlin: Springer; 2000. [24] Maki KL, Braun RJ, Ucciferro P, Henshaw WD, King-Smith PE. Tear film dynamics on an eye-shaped domain. Part 2. Flux boundary conditions. J Fluid Mech 2010;647:361–90. [25] Maki KL, Braun RJ, Driscoll TA, King-Smith PE. An overset grid method for the study of reflex tearing. Math Med Biol 2008;25:187–214. [26] Pult H, Korb D, Murphy P, Blackie C, Riede-Pult B. Central lid margin position and tear film spreading during ‘lid closure’ in spontaneous blinking. In: 7th international conference on the tear film & ocular surface: basic science and clinical relevance. 2013. [27] Yokoi N, Yamada H, Mizukusa Y, Bron AJ, Tiffany JM, Kato T, et al. Rheology of tear film lipid layer spread in normal and aqueous tear – deficient dry eyes. Investig Ophthalmol Vis Sci 2008;49:5319–24. [28] Shen M, Li J, Wang J, Ma H, Cai C, Tao A, et al. Upper and lower tear menisci in the diagnosis of dry Eye. Investig Ophthalmol Vis Sci 2009;50:2722–6. [29] Wang J, Aquavella J, Palakuru J, Chung S, Feng C. Relationships between central tear film thickness and tear menisci of the upper and lower eyelids. Investig Ophthalmol Vis Sci 2006;47:4349–55. [30] Yang L, Tu Y, Fang H. Modeling the rupture of a capillary liquid bridge between a sphere and plane. Soft Matter 2010;6:6178–82. [31] Berry M, Pult H, Purslow C, Murphy PJ. Mucins and ocular signs in symptomatic and asymptomatic contact lens wear. Optom Vis Sci 2008;85:E930–8. [32] Pult H, Purslow C, Berry M, Murphy PJ. Clinical tests for successful contact lens wear: relationship and predictive potential. Optom Vis Sci 2008;85:E924–9. [33] Pult H, Purslow C, Murphy PJ. The relationship between clinical signs and dry eye symptoms. Eye (Lond) 2011;25:502–10. [34] Höh H, Schirra F, Kienecker C, Ruprecht KW. Lid-parallel conjunctival folds are a sure diagnostic sign of dry eye. Ophthalmologe 1995;92:802–8. [35] Pult H, Riede-Pult B. Impact of lid-parallel conjunctival folds on the tear meniscus thickness. In: BCLA 2014. 2014.
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
ARTICLE IN PRESS
CLAE-772; No. of Pages 8
Contact Lens & Anterior Eye xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Contact Lens & Anterior Eye journal homepage: www.elsevier.com/locate/clae
A new model of central lid margin apposition and tear film mixing in spontaneous blinking Heiko Pult a,b,∗ , Donald R. Korb c , Paul J. Murphy d , Britta H. Riede-Pult a,b , Caroline Blackie c a
Dr. Heiko Pult – Optometry & Vision Research, Germany Cardiff University, School of Optometry & Vision Sciences, UK c Korb Associates and TearScience Inc., USA d University of Waterloo, School of Optometry & Vision Science, Canada b
a r t i c l e
i n f o
Article history: Received 13 August 2014 Received in revised form 21 January 2015 Accepted 26 January 2015 Keywords: Over-blink Spontaneous blink Tear film Tear film spreading Central lid position Spontaneous blinks Dry eye
a b s t r a c t Purpose: To investigate tear film spreading and central lid position in spontaneous blinks and to propose a model of the central lid position of the closed eye in such blinks. Method: In vivo: lid margin position and geometry of 15 subjects (9 female; median age = 45 years) were evaluated by high-speed video and slit-lamp microscope video in consecutive spontaneous blinks. Upper lid (UL) tear meniscus (TM) depth was observed in the open and almost closed eye. Eyelid geometry, position and UL TM depth were analysed by Image-J Software. Lid margin thicknesses were measured with a Scheimpflug camera. In vitro: tear film spreading and lipid layer formation were simulated on a lubricated glass plate and videoed by high-speed camera (JVC, GZ-GX1BE, Japan). Results: In vivo: the median central lid margin thickness was not significantly (p = 0.258) different between UL (1.8 mm) and LL (1.7 mm) in the opened eye. During blinking, UL remained perpendicular to the corneal surface, while LL tilted in and thinned. A scaled model diagram was created and revealed an over-blink of the UL over the LL (>0.7 mm) and a height offset of the posterior lid margin of >0.7 mm. In vitro: the LL TM fused with the UL TM even before full lid touch due to capillary bridge building. Conclusions: The central UL overlaps the central LL during spontaneous blinking. This provides the appearance of complete closure. The space that results from the lack of lid margin apposition influences the fusion of the upper and lower TM and ultimately tear film mixing. © 2015 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved.
When a human eye opens fully, in primary gaze, the upper eyelid margin is moved upwards and the surface of the margin is angled forwards, while the lower lid margin is moved downwards and the margin surface angled forwards [1]. During eyelid closure, the orbicularis muscle pulls the upper lid down towards the lower lid and the lower lid is pulled slightly nasally [2,3]. Bron et al. [4] state that during lid closure the occlusal surfaces of the lid margins are apposed. There appears to be an inherent assumption in our understanding of the blink that during lid closure, the lid margins, specifically the occlusal surfaces, come into contact [5,6]. This apposition results in fusion of the tear menisci and lipid reservoirs along both lid margins [7,8]. In the inversion of this argument, any fusion of the tear menisci and lipid reservoirs along the lid margins
∗ Corresponding author at: Dr. Heiko Pult – Optometry & Vision Research, Germany. Tel.: +49 1749025090; fax: +49 6201873657. E-mail address: [email protected] (H. Pult).
would not be possible without sufficient lid touch. Based on this assumption, it was hypothesised that there must be some rotation of the upper lid margin surfaces during blinking to bring the lid margins into apposition [1]. However, recent evidence has demonstrated that the central portion of the upper and lower lid margins frequently do not touch during spontaneous blinking [5,9]. Further, an analysis of highspeed video recording of the blinks demonstrated an over-lapping of the central upper lid over the central lower lid, named as an ‘overblink’. This over-blink likely results in a space posterior to the inner upper lid as the central upper lid margin rests over the lower lid lashes. It has been suggested that the over-blink is arguably the normal state in healthy Caucasians and a possible reason for the observation of an unaltered drop of vital dye placed on the lower lid margin during blinking [9]. These observations raise two questions: (i) what positions do the lid margins actually take in a completed, spontaneous blink; and
http://dx.doi.org/10.1016/j.clae.2015.01.012 1367-0484/© 2015 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
G Model CLAE-772; No. of Pages 8
ARTICLE IN PRESS H. Pult et al. / Contact Lens & Anterior Eye xxx (2015) xxx–xxx
2
(ii) how does tear film mixing occur, leading to the establishment of a proper lipid layer. Since the over-blink is likely the normal state in healthy Caucasians, the need for full lid touch during a blink for proper tear film mixing is called into question. We hypothesise that even within this small space between central the lids, both tear menisci can fuse during blink lid closure due to rheological mechanisms. However, to better understand these mechanisms, more detailed information about the central lid position of the closed eye in spontaneous blinks is needed. Furthermore, possible rheological mechanisms behind this new blink model need to be investigated. Having answered these questions may give us a better understanding of tear film mixing and lipid layer formation in the over-blink lid position, since the anecdotal concept of tear menisci fusion based on the lid touch may not be relevant. The purpose of this pilot-study was to investigate these questions by examining tear film spreading and central lid position of the closed eye in spontaneous blinks in a cohort of human subjects, and to propose an in vitro model of the central lid position of the closed eye in complete, spontaneous blinks.
Fig. 1. Sketch showing the different camera positions (1: temporal view; 2: inferiortemporal view; 3: frontal view).
1. Method
1.2. Techniques (Table 1)
For the human study, 15 healthy subjects (9 females; median age = 45 years) were recruited from the clinical practice Horst Riede GmbH, Weinheim, Germany. As has been previously described elsewhere, a 0.15 l lissamine green drop was placed on the anterior portion of the central lower lid margin of the lower lid (LL) [5,9]. Each subject was asked to blink normally for 30 s, and lid margin position and geometry were evaluated by high-speed video and slit-lamp microscope video. UL tear meniscus depth (z-axis) was observed in the completely opened eye and almost closed eye using slit-lamp photography. The term ‘almost closed’ describes the smallest possible lid aperture in which the lid margins still were observable. In this position the UL and LL margins still show a vertical gap of up to half of the UL margin’s thickness (from the temporal-inferior view) [10]. Eyelid geometry, margin position and upper lid tear meniscus depth were analysed by Image-J Software (Version 1.42q; Wayne Rasband, National Institute of Health, Bethesda, MD, USA) from images taken from the video. Lid margin thicknesses and corneal topography were measured with a Scheimpflug camera (Pentacam, Oculus Optikgeräte GmbH, Wetzlar, Germany). Based on these data, a scaled model diagram was created to represent the central lid position of the closed eye in spontaneous blinking. All observations were done by the same observer (HP).
1.2.1. Lissamine green drop test This test was used to exclude touch of the central lids in spontaneous, complete blinks and to evaluate the minimum distance between the UL and LL margins in blinks. Subjects were positioned on the chin-rest of the slit-lamp microscope and a 0.15 l lissamine green drop was placed onto the anterior portion of the central, LL using a microliter pipette (neoLab-Mikroliterpipette Economy, neoLab Migge Laborbedarf-Vertriebs GmbH, Heidelberg, Germany). The anterior portion was defined as extending from the meibomian gland orifices to the eyelash line [5]. The lissamine green drop was photographed at each subject’s lid margin and the drop height was measured using Image-J software. 1.2.2. Video Subjects were asked to maintain normal spontaneous blinking over a 30 s period while looking straight ahead. A frontal view of blinking was digital recorded using a video slit-lamp microscope (DigiPro2, bon Optic, Lübeck, Germany). Simultaneously, the same blinks were videoed from a temporal-inferior view (Fig. 1) using a high-speed video camera (250 frames/s; JVC, GZ-GX1BE, Japan). A previous investigation revealed that the temporal-inferior view provided the best observation of the over-blink [9]. The extent of over-blink was analysed subjectively on a 5-grade scale. Additionally, blinks were videoed at high-speed from a side-view to analyse lower lid movement. Number of completed, spontaneous blinks were recorded; any potential non-spontaneous blink was excluded by observing Bell’s phenomenon.
1.1. Exclusion criteria Subjects were excluded if they had floppy eyelids [11] (easily everted, floppy upper eyelid), conjunctivo chalasis, irregular lid margins (such as irregularity in, for example, cicatricial meibomian gland dysfunction), supra-nuclear motor impairment of lid movements, ptosis, apraxia, rheumatoid arthritis, diabetes, recent ocular infections, hay fever, any history of ocular surgery, use of any medication or eye drops known to affect the ocular surface, worn contact lenses or had contact lens experiences, or were pregnant. All procedures were conducted in accordance with the Declaration of Helsinki (2000), and approval for the study was given by the Cardiff School of Optometry and Vision Sciences Ethics Committee. All subjects were fully consented prior to study enrolment.
1.2.3. Geometry and position of the central eye lids Upper eyelid: the relative position of the UL to the cornea was measured from the high-speed video recordings using Image-J software. Imaging of the lid margins of the opened eye were captured by Scheimpflug camera (Pentacam, Oculus, Wetzlar, Germany) and central lid margin thickness (distance between eye lash line and cornea) and angle of the lid margin surface, relative to the cornea, were measured from the images using the analysis tool of the Pentacam software. Variations of the UL margin thickness during each blink were observed by measuring the distance between the eyelash line and the UL tear meniscus, as well as the distance between the eyelash line and cornea. The position of the UL were photographed in the fully opened eye and almost closed eye using the slit lamp microscope, assisted by the use of a small dental mirror
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
G Model CLAE-772; No. of Pages 8
ARTICLE IN PRESS H. Pult et al. / Contact Lens & Anterior Eye xxx (2015) xxx–xxx
3
Table 1 Summarising table of the different techniques. Technique
Purpose
Procedure
(i) To exclude touch of the lid margins in complete, spontaneous blinks (ii) To analyse the minimal distance between central lid margins after completed, spontaneous blink
(i) 0.15 l lissamine green drop placed onto the anterior portion of the central, LL (ii) Measurement of the lissamine green drop height
To analyse central lower lid vertical movement, eye ball retraction and bells phenomenon
High-speed video from the temporal view (camera position 1) and digital image analyses
(i) To classify the over-blink (ii) To analyse the relative position of the UL to the cornea in complete, spontaneous blinks
High-speed video from the inferior-temporal view (camera position 2) and subjective and digital image analyses
Frontal view photography
To analyse lissamine green drop height
Digital slit lamp microscope observation (camera position 3) and digital image analyses
Inferior view photography
To measure variations of the UL in the opened and almost closed position
Digital slit lamp microscope plus dental mirror to observe the UL margin and digital image analyses
Scheimpflug camera
To analyse central lid margin thickness and relative eye lid position to the cornea of the opened eye
Scheimpflug camera observation and digital image analyses
Glass plate model
To simulate tear film spreading and tear meniscus fusion
Glass plate model with saline solution and different coloured dyes. High-speed videoing and digital image analyses
Lissamine green drop test
Video: side view
Video: inferior-temporal view
positioned to permit viewing of the eyelid margin from below for the almost closed eye. The mirror helped to avoid any perspective error that would make the almost closed UL appear thicker when viewed from the front. Lower eyelid: to measure the LL position, a small horizontal slit beam from the slit lamp illumination was projected onto the central LL margin, directly below the centre of the subject’s pupil. Since full closure of the UL would hinder the view of the LL margin, the subject’s eyebrow was gently lifted by the patient during measurements. Although only a slight movement on the eyelid was produced, this was sufficient to assist in viewing of the LL, while not interfering with normal lid movement. Blinks were again highspeed videoed from a side view, and central LL vertical movement were measured using the position of the projected small horizontal light slit with Image-J software. Finally, any retraction of the eyeball during blinks was analysed by Image-J software by observing the relative movement of the eyeball.
1.2.4. Over-blink The over-blink of the UL during complete blinks was evaluated by analysing the high-speed video (inferior-temporal view) and classified subjectively (Fig. 2) using a 5-grade scale [9] (0: aligned; 1: over-blink of UL extending less than 1/3 of UL margin thickness, 2: over-blink of UL extending between 1/3 and 2/3 of UL margin thickness, 3: over-blink of UL extending between 2/3 and 3/3 of UL margin thickness, 4: over-blink of UL extending more than 3/3 of UL margin thickness).
Fig. 2. Grading scale of the over-blink [9].
1.2.5. Upper lid tear meniscus At least 30 min after observations described above, the depth (zaxis) of the UL tear meniscus was photographed in the fully opened eye lid position and almost closed eye lid position using the slit lamp microscope and a small dental mirror. The tear meniscus was coloured by fluorescein to assist in observation. To allow repeatable fluorescein application, the fluorescein strip (Bio Glo, Biovision Ltd., Dunstable, UK) was modified according to Pult et al. [12]. In summary, a normal fluorescein strip was modified by folding the dye impregnated tip at the 1 mm length (named the modified fluorescein strip (MFS)), which was then moistened with a single drop of saline (Software Saline, Alcon, Großostheim, Germany) and shaken to remove excess fluorescein before instillation. 1.2.6. Glass plate model For the in vitro model study, tear film spreading and lipid layer formation were simulated on a glass plate model lubricated with saline solution (Software Saline, Alcon, Großostheim, Germany; concentration: 0.9% (based on manufacturer’s data)) (Figs. 3 and 4). Two glass plates, representing the lids, with a thickness of 1.9 mm each, were placed on a glass plate base. The thickness of the glass plates was measured using a calliper rule. The base plate was loaded with the saline solution and a small lissamine green drop was placed at the tear meniscus of the glass plate representing the lower lid. Blinks without ‘lid-touch’ were simulated by moving the second glass plate (upper lid) to and from this first glass plate (lower lid), without touching, and videoed by high-speed camera. The model was then observed to see, if after at least one simulated “noncontacting” blink, the lissamine green had spread across from the upper lid to the lower lid glass plates. To then simulate mixing of the lipid layer, red coloured oil (St. John’s Wort red oil, Dr. L. Schmitt GmbH & Co. KG, Planegg, Germany) was placed on top of the saline layer, along with a small green oil drop (pumpkin seed oil, Fausere Vitaquellwerk KG, Hamburg, Germany) placed at the tear meniscus of the lower eyelid glass plate. Potential mixing of the tear film was then evaluated subjectively by simulating blinks using movement of the glass plates. In addition to subjective evaluation, the glass plate model was videoed from a side view and loaded with the minimum volume of saline solution to achieve full wetting of the glass plate base
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
G Model CLAE-772; No. of Pages 8
ARTICLE IN PRESS H. Pult et al. / Contact Lens & Anterior Eye xxx (2015) xxx–xxx
4
Fig. 3. Simulation of aqueous tear film mixing in the glass model (lubricant: saline solution; one tear meniscus was coloured by lissamine green). Left: simulated closing phase of a blink. Middle: completed closing phase without ‘lid-touch’ (=return point). Right: opening phase of the simulated blink. Lissamine green was taken from the lower lid glass plate and is distributed over the lid aperture and is already in the upper lid glass plate tear meniscus.
Fig. 4. Simulation of lipid layer mixing in the glass model. (A) Model is lubricated by saline solution. Small green oil drop (black arrow) at the lower lid glass plate meniscus and red oil spread over the total ‘tear-film’ (dashed arrow). (B) Simulated closing phase of a blink. (C) Opening phase of the simulated blink. Green oil drop is slightly mixed with the red oil (dashed arrows).
and to produce visible tear menisci along the glass edges of both overlying plates. A subjective assessment was then made whether tear meniscus fusion would occur, depending on the minimum and maximum possible saline load of the model. The glass plate experiments were trained several times before the experiment, especially to guarantee repeatability and appropriate simulated blink speed. In the experiment the achieved simulated blink speed and the distance until first tear menisci fusion was analysed from the videos of 5 consecutive experiments. 1.2.7. Statistical analyses Data were examined for normality by Shapiro–Wilk test. Correlations between measurements and the over-blink (ordinal scale, non-parametric) were evaluated by Spearman rank. Differences between lid thicknesses (parametric) were evaluated by paired t-test. The data were analysed using SPSS 20.0 (SPSS Inc., Chicago, USA) and WinStat 2007.1 (R Fitch Software, Bad Krozingen, Germany). 2. Results 2.1. Human studies The lissamine green drop was observed to remain unaltered for a median of 8 consecutive, complete blinks for all subjects (Table 2). The observed median over-blink grade was 3 (interquartiles: 3–4). The eyeball was drawn back into the orbit during blinking by a median distance of 0.9 mm (interquartiles: 0.36–0.97). The median height of the instilled lissamine green drop at subject lid margins was 0.4 mm (y-axis; interquartiles: 0.37–0.47) and such slightly higher than expected of a 0.15 l drop. Since the drop was not altered after consecutive blinks, the vertical space between the
upper and lower central anterior lid margins must have been at least 0.4 mm in the closed eyelid position. The median central lid margin thickness was slightly, but not significantly (p = 0.258), larger for the upper lid (1.8 mm) than for the lower lid (1.7 mm) when the eyelids were open. The upper lid margin was also observed to remain perpendicular to the corneal surface at all times during blinking. The lower lid margin tilt changed from the fully open position by 23◦ (median, upwards, inwards) and thinned by 20.0% (Table 2). This change in thickness was significantly correlated to the over-blink score (r = 0.791; p = 0.0032). The distance between the upper lid eyelash line and cornea also increased by 1.17 times (median) from the fully opened position to the almost closed position. This increase was significantly correlated to overblink scores (r = 0.779; p = 0.002). Therefore the median distance between cornea and eyelash line of the upper lid appears to be larger in the closed eyelid position. Concurrently, the distance between the upper lid eyelash line and the outer margin of the upper lid tear meniscus only enlarged by 1.09 times (median), between the fully opened position and the almost closed position. This parameter was significantly correlated to over-blink scores (r = 0.553; p = 0.039). At the same time, the tear meniscus depth increased by a median ratio of 1.8, which was also significantly correlated to over-blink scores (r = 0.875; p < 0.001). In other words, in the almost closed eye, the upper eyelid appears to thicken and move away from the corneal surface (cease to touch the cornea), with the space created being filled by the tear film, revealed by an expanded tear meniscus depth. To further investigate these changes, a scale diagram of the lid positions was constructed from the evaluated data (Fig. 5). Observation revealed an over-blink of the UL over the LL by >0.7 mm (z-axis) and a height offset of the posterior lid margin of >0.7 mm
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
G Model CLAE-772; No. of Pages 8
ARTICLE IN PRESS H. Pult et al. / Contact Lens & Anterior Eye xxx (2015) xxx–xxx
5
Table 2 Median values and interquartiles of the observed parameters. Means and standard deviation (SD) are additionally shown in parametric data, only. Parameter
Median
In vivo experiment Number of consecutive spontaneous complete blinks Draw back of the eye ball [mm] Lissamine green drop height [mm] Over blink Upper lid thickness (opened) [mm] Upper lid thickness (closed) [mm] Lower lid thickness (opened) [mm] Lower lid thickness (closed) [mm]
8 0.9 0.4 3 1.8 2.1 1.7 1.4
In vitro experiment Distance between glass plates at tear menisci fusion [mm] Simulated glass plate speed [cm/s]
0.5 23
Mean
±SD
7–13 0.35–0.97 0.37–0.47 3–4 1.56–1.90 1.65–2.25 1.56–1.90 1.23–1.55
9 0.7 0.4 n/a 1.7 2.0 1.7 1.4
3.79 0.14 0.12 n/a 0.22 0.25 0.24 0.23
0.35–0.65 15.8–23.0
0.5 20
0.16 3.92
Interquartiles
Fig. 6. Simulation of the menisci fusion without lid touch (saline solution, red vertical mark on glass plates = reference mark: 5 mm increments, dashed lines show tear meniscus height). Top: tear meniscus formation at both glass plates. Right glass plate is moved towards the left one. Middle: capillary bridge formation even distance between tear menisci was still 0.23 mm. Below: very thick capillary bridge when distance between tear menisci was zero.
Fig. 5. Scale diagram of lid position in the fully opened eye and closed eyelid positions, during spontaneous, complete blinks (median values are shown).
(y-axis), which produced a triangular shaped space between the cornea and lid margins. 2.2. In vitro The LL tear film meniscus was observed to fuse with the UL tear meniscus (Figs. 3 and 4) even before full lid touch (median distance between the glass plates was 0.5 mm in a simulated complete blink). Similarly, lipid spread occurred without any ‘lid-touch’, i.e. no contamination of the lower lid glass plate with coloured fluid from the UL glass plate during the opening phase. Simulated blink
speed was 23 cm/s (median speed of 5 consecutive experiment setups (Table 2)), which is close to reported figures in humans (20 cm/s [13]). To observe tear film fusion from a side view, the UL glass plate was moved towards the LL glass plate (∼5 cm/s, Fig. 6). This demonstrated that both tear menisci quickly build up a capillary bridge (Fig. 6). Subjectively, this fusion of the tear menisci occurred earlier (at a larger distance between plates) with maximum saline loading compared to minimal loading of the glass plates. Furthermore, it appeared that capillary bridge building developed shortly before the tear menisci outer limits met. Thus, for the tear menisci to merge, the lids do not need to meet, but just approach by a certain distance (0.23 mm in the glass plate model, Fig. 4). For irregular glass plate margins, multiple capillary bridges occurred, which enabled partial tear film mixing (Fig. 7), depending on the tear film volume (Fig. 8). The resting time of the plates in close proximity to each other also seemed to affect total tear film mixing, presumably due to changes in surface tension across the capillary bridge, and subsequent capillary flow.
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
G Model CLAE-772; No. of Pages 8
ARTICLE IN PRESS H. Pult et al. / Contact Lens & Anterior Eye xxx (2015) xxx–xxx
6
Fig. 7. Left image: initial development of capillary bridges between the two glass plates at three locations (right glass plate lubricant: fluorescein in saline solution; left plate: saline solution) showing partial tear film mixing. Right image: Simulated opening phase of blink.
3. Discussion This study has confirmed the results of previous investigations in Caucasians [5,9] and found that a drop of vital dye placed on the central lower eyelid margin remained unchanged after several (median, 8) consecutive, apparently complete, spontaneous blinks. An over-blink was also observed, with a median subjective grade of 3. This was slightly smaller than the expected over-blink calculated from the scaled model diagram. It was also observed that, while the UL margin swept normal to the corneal surface during blinking, the margin also thickened as it approached the LL margin for apparent closure. As the LL rotated inwards and moved horizontally towards the nose, the margin itself also thinned slightly by 0.3 mm (median) or 20%. The lissamine green drop with an median height of 0.4 mm was unaltered after normal spontaneous blinks, it can be assumed that the central lid margins of the UL and LL do not make direct contact, and that the minimum distance between the UL and LL is larger than the lissamine drop height. These changes in lid thickness during blinking, along with the UL over-blink placement, resulted in a small space between the posterior, central, UL near the margin and the corneal surface of at least 0.7mm in this cohort of patients. Furthermore, the scale model diagram illustrates that the lids appear to be fully closed from a frontal view, as commonly observed in humans [14–17]. The LL thinning is likely the result of the inward, nasal movement of the eyelid during spontaneous blinking [18,19]. However, the apparent thickening of the UL margin during blinking appears to be new information. This phenomenon could be a relative weakness of the orbicularis oculi muscle, particularly of the pretarsal portion, resulting in a slight lifting off of the lid wiper away from the cornea [9], or, more likely, due to muscle thickening during muscle contraction of the pretarsal portion of the orbicularis oculi muscle pulling the eyelid away from the cornea. There may also be an effect from the modulus of elasticity of the lid wiper portion of the eyelid, which is compressed in the fully opened eye, but becomes more relaxed during a blink [20]. Although the study conditions necessary for observation meant that these changes were seen when the subject was asked to make a voluntary blink to achieve an almost closed eye, it seems reasonable to assume that this situation will also occur during spontaneous blinking. The increased distance between the eyelash line and cornea, and the increased tear meniscus, support the hypothesis that the posterior eyelid margin contour changes during a blink. While Kessing’s space [21], between the tarsal UL and bulbar conjunctiva, has long been recognized, little is known regarding the dynamics of this space during blinking. Specifically, changes to the lid wiper contour
and the interaction between the cornea and lid wiper during blinking have not been previously documented [20,22]. One possible hypothesis is that the modulus of the lid wiper is promoting the enlargement of the distance between cornea and anterior lid margin at the almost complete eye lid position compared to the opened one, since lid pressure may be different between both positions [20]. Also, although the central UL margin remains perpendicular to the cornea throughout the blink, its final position in the closed eye appears to be angled inwards by approximately 64◦ , relative to a normal to the centre of the cornea. Similarly, the central lower lid tilts inwards by approximately 23◦ , relative to the normal. To our knowledge this is the first time that the angular position of both central eyelid margins has been reported in spontaneous blinks. However, the measurements used to produce the scale model diagram are limited by the available techniques. To our knowledge there is no technique available that allows the observation of lid position in spontaneous blinks non-invasively and with a high-speed sectional imaging. The scaled diagram of this pilotstudy is also based on a small population of mid-aged Caucasians and it would be of interest whether there is a difference between gender, races, or between symptomatic and asymptomatic dry eye patients in a larger population. Nevertheless, observations are well correlated to subjective over-blink scores showing that the scaled diagram approximates what really happens in such observed blinks. Therefore the triangular space between the closed eyelids is of significant interest and deserves continuing investigation. The model may also illustrate some practical benefits from not having full, parallel touch of the lids – for example, if it did occur, the tear film may be squeezed out, the meibomian gland orifices may block each other, or there may not be sufficient space for the lipid layer to compress into with a full blink. Since the results of the in vivo experiment demonstrated that the final tear film characteristics are not dependent on complete central lid margin apposition, tear film mixing was evaluated by using glass plates to simulate blinking, lubricated by saline solution. This in vitro experiment demonstrated that the two tear menisci do not need to overlap to mix. This is possibly due to the formation of a capillary bridge. The side view of the glass plates demonstrated this capillary bridge between the UL glass plate tear meniscus and the LL glass plate tear meniscus. The energy of capillary forces in fluid rheology are described as being very high, mainly depending on the lubricant surface tension [23], making it a potentially important mechanism in tear film mixing, at least for the glass plate blink model. To create a capillary bridge, both tear menisci must be gradually moved closer together, until they reach a certain critical distance
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
G Model CLAE-772; No. of Pages 8
ARTICLE IN PRESS H. Pult et al. / Contact Lens & Anterior Eye xxx (2015) xxx–xxx
7
Fig. 8. Difference in capillary bridge building produced by varying saline solution loading of glass plates; lower load (left) in comparison to higher load (right). The white arrows show likely flux of the fluorescein from beneath the glass plates towards space between plates (lid aperture).
apart. In the glass plate model it appeared that the extending edges of each tear meniscus do not even need to touch (Fig. 6) for capillary bridge building to occur shortly before touch. Capillary bridge building was related to loading of the model, therefore the tear film volume and lid geometry may have a great impact in spontaneous blinks. After any blink, the capillary bridge will collapse in the opening phase once a certain distance has been created between the lid margins and tear flux [24,25] will draw the tear film into the tear film menisci. Since central lid margins do not commonly touch during a blink and tear film volume appears to impact tear meniscus mixing in the closed eyelid position in spontaneous blinks, tear film volume and eye lid position in such blinks may be vital for the formation of a capillary bridge [26]. However, even if central tear menisci are not brought close enough for full bridging to occur, para-central capillary bridges may still be able to promote sufficient tear film mixing (Figs. 7 and 8). The glass plate model of the in vitro experiment is an approximation of the real eye, especially in terms of tear film viscosity, ocular surface topography and eye lid tension. However, the in vivo observation of the para-central capillary bridge building (Fig. 9) may demonstrate that the rheological mechanisms observed by the in vitro experiment are transferable to the real eye. It has been previously noted that tear volume may play a role in promoting lipid layer formation [27], thus, it is possible that this bridging phenomenon assists in the uptake of lipids from the lipid reservoir. Assuming a normal tear meniscus height range (in the open eye) for the UL of 0.17 mm [28] to 0.27 mm [29] and LL of 0.20 mm [28] and 0.31 mm [29], and comparing it to the distance between the posterior lid margins, modelled as being at least 0.7 mm, the distance between the two tear menisci would seem to be too large to allow for capillary bridging. However, observation of the UL tear meniscus in the almost closed eye showed that the meniscus enlarges, probably due to the tear film thickening as the same volume covers a smaller ocular surface area. This will increase tear volume along the lid margins and cause the tear menisci to extend further away from the margins in the very first moment of the completed blink, thereby reducing the distance between the two menisci and permitting capillary bridging. A space between the lid margins and the cornea would facilitate this, just as what appears to occur with the over-blink. Further investigation of these effects
Fig. 9. In vivo simulation of a capillary bridge between temporal tear menisci coloured by fluorescein. Top: early stage of capillary bridge formation (white arrow). Middle: full para-central capillary bridge (dashed arrow) after further minimal reduction of lid aperture. Below: sketch of full capillary bridge. Upper tear meniscus height next to the capillary bridge was 0.18 mm, the lower lid tear meniscus height was 0.17 mm and the distance between both was 0.30 mm.
in dry eye patients and contact lens wearers would be of great interest. The capillary force due to a liquid bridge, so called capillary bridge, and the breaking of this bridge, are of high importance due to the widespread existence of capillary menisci in macro-, micro-, and even nano-scale applications [30]. The capillary bridge facilitates the fusion of both tear menisci even if the central UL and LL tear menisci are not overlapping, as is probably the case in the over-blink. Until now, tear film fusion was considered to be possible only with lid touch in blinks. This appeared to be especially important in lipid layer formation [7,8]. In contrast, the over-blink is assumed to be the normal state in healthy Caucasians [9]. Furthermore, the investigation of this paper had shown that there appears to be a noticeable space between the posterior lid margins, and
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
G Model CLAE-772; No. of Pages 8
ARTICLE IN PRESS H. Pult et al. / Contact Lens & Anterior Eye xxx (2015) xxx–xxx
8
consequently between the UL and LL tear menisci. All of this calls the importance of the lid touch in tear film mixing into question. The observation of the capillary bridge formation may be the missing piece of information in tear film fusion in spontaneous, complete blinks (Fig. 9). This model for tear mixing using capillary bridging in the space created by the over-blink might be altered if the capillary bridging was impeded in some way, perhaps by the presence of lid parallel conjunctival folds (LIPCOF). LIPCOF scores are increased in symptomatic dry eye patients, and they are correlated to lidwiper epitheliopathy [31–34]. LIPCOF may physically hinder the blink action as it approaches eyelid closure [10] and they also influence tear meniscus distribution along the lid margin [35]. Therefore in patients with an increased LIPCOF score, insufficient para-central capillary bridge formation is likely, and tear film mixing, spreading and lubrication of the bulbar conjunctiva may be insufficient. This effect may only be dominant in the opening phase of a blink, but this phase may be vital for lipid layer spreading. Such an effect remains to be confirmed. 4. Conclusions The results of this study show that, in contrast to prior assumptions regarding lid apposition during blinking, the central upper lid margin overlaps the central lower lid margin during spontaneous blinking. This overlap of the upper lid margin provides the appearance of complete closure. The space that results from the lack of lid margin apposition influences the fusion of the upper and lower menisci and, ultimately, tear film mixing. Financial support None. Conflict of interest No conflicting relationship exists for any author. References [1] Foulks GN, Bron AJ. Meibomian gland dysfunction: a clinical scheme for description, diagnosis, classification, and grading. Ocul Surf 2003;1:107–26. [2] Bell C, Bell C. On the motion of the eye, an illustration of the uses of the muscles and nerves of the orbit. Philos Trans R Soc Lond 1823;113:166–86. [3] Doane MG. Blinking and the mechanics of the lacrimal drainage system. Ophthalmology 1981;88:844–51. [4] Bron AJ, Yokoi N, Gaffney EA, Tiffany JM. A solute gradient in the tear meniscus. I. A hypothesis to explain Marx’s line. Ocul Surf 2011;9:70–91. [5] Korb DR, Blackie CA, McNally E. Evidence that the upper and lower lids do not make complete contact even when the lids are closed. Cornea 2013;32:491–5. [6] Bron AJ, Yokoi N, Gaffney EA, Tiffany JM. A solute gradient in the tear meniscus. II. Implications for lid margin disease, including meibomian gland dysfunction. Ocul Surf 2011;9:92–7.
[7] Bron AJ, Tiffany JM. The meibomian glands and tear film lipids. Structure, function, and control. Adv Exp Med Biol 1998;438:281–95. [8] Bron AJ, Tiffany JM, Gouveia SM, Yokoi N, Voon LW. Functional aspects of the tear film lipid layer. Exp Eye Res 2004;78:347–60. [9] Pult H, Riede-Pult B, Murphy PJ. A new perspective on spontaneous blinks. Ophthalmology 2013;120:1086–91. [10] Pult H, Riede-Pult BH, Murphy PJ. The relation between blinking and conjunctival folds and dry eye symptoms. Optom Vis Sci 2013;90:1034–9. [11] Culbertson WW, Ostler HB. The floppy eyelid syndrome. Am J Ophthalmol 1981;92:568–75. [12] Pult H, Riede-Pult BH. A new modified fluorescein strip: its repeatability and usefulness in tear film break-up time analysis. Contact Lens Anterior Eye 2012;35:35–8. [13] Doane MG. Interactions of eyelids and tears in corneal wetting and the dynamics of the normal human eyeblink. Am J Ophthalmol 1980;89:507–16. [14] Doughty MJ. Further assessment of gender- and blink pattern-related differences in the spontaneous eyeblink activity in primary gaze in young adult humans. Optom Vis Sci 2002;79:439–47. [15] Doughty MJ. Consideration of three types of spontaneous eyeblink activity in normal humans: during reading and video display terminal use, in primary gaze, and while in conversation. Optom Vis Sci 2001;78:712–25. [16] Bentivoglio AR, Bressman SB, Cassetta E, Carretta D, Tonali P, Albanese A. Analysis of blink rate patterns in normal subjects. Mov Disord 1997;12:1028–34. [17] Collins M, Heron H, Larsen R, Lindner R. Blinking patterns in soft contact lens wearers can be altered with training. Am J Optom Physiol Opt 1987;64: 100–3. [18] Casse G, Sauvage JP, Adenis JP, Robert PY. Videonystagmography to assess blinking. Graefe’s Arch Clin Exp Ophthalmol 2007;245:1789–96. [19] Spooner J. Ocular anatomy. London: The Hatton Press Ltd.; 1957. [20] Jones M, Fulford G, Please C, McElwain D, Collins M. Elastohydrodynamics of the eyelid wiper. Bull Math Biol 2008;70:323–43. [21] Kessing SV. A new division of the conjunctiva on the basis of X-ray examination. Acta Ophthalmol (Copenh) 1967;45:680–3. [22] Dunn AC, Uruena JM, Huo Y, Perry SS, Angelini TE, Sawyer WG. Lubricity of surface hydrogel layers. Tribol Lett 2013;49:371–8. [23] Persson BNJ. Sliding friction. 2nd ed. Berlin: Springer; 2000. [24] Maki KL, Braun RJ, Ucciferro P, Henshaw WD, King-Smith PE. Tear film dynamics on an eye-shaped domain. Part 2. Flux boundary conditions. J Fluid Mech 2010;647:361–90. [25] Maki KL, Braun RJ, Driscoll TA, King-Smith PE. An overset grid method for the study of reflex tearing. Math Med Biol 2008;25:187–214. [26] Pult H, Korb D, Murphy P, Blackie C, Riede-Pult B. Central lid margin position and tear film spreading during ‘lid closure’ in spontaneous blinking. In: 7th international conference on the tear film & ocular surface: basic science and clinical relevance. 2013. [27] Yokoi N, Yamada H, Mizukusa Y, Bron AJ, Tiffany JM, Kato T, et al. Rheology of tear film lipid layer spread in normal and aqueous tear – deficient dry eyes. Investig Ophthalmol Vis Sci 2008;49:5319–24. [28] Shen M, Li J, Wang J, Ma H, Cai C, Tao A, et al. Upper and lower tear menisci in the diagnosis of dry Eye. Investig Ophthalmol Vis Sci 2009;50:2722–6. [29] Wang J, Aquavella J, Palakuru J, Chung S, Feng C. Relationships between central tear film thickness and tear menisci of the upper and lower eyelids. Investig Ophthalmol Vis Sci 2006;47:4349–55. [30] Yang L, Tu Y, Fang H. Modeling the rupture of a capillary liquid bridge between a sphere and plane. Soft Matter 2010;6:6178–82. [31] Berry M, Pult H, Purslow C, Murphy PJ. Mucins and ocular signs in symptomatic and asymptomatic contact lens wear. Optom Vis Sci 2008;85:E930–8. [32] Pult H, Purslow C, Berry M, Murphy PJ. Clinical tests for successful contact lens wear: relationship and predictive potential. Optom Vis Sci 2008;85:E924–9. [33] Pult H, Purslow C, Murphy PJ. The relationship between clinical signs and dry eye symptoms. Eye (Lond) 2011;25:502–10. [34] Höh H, Schirra F, Kienecker C, Ruprecht KW. Lid-parallel conjunctival folds are a sure diagnostic sign of dry eye. Ophthalmologe 1995;92:802–8. [35] Pult H, Riede-Pult B. Impact of lid-parallel conjunctival folds on the tear meniscus thickness. In: BCLA 2014. 2014.
Please cite this article in press as: Pult H, et al. A new model of central lid margin apposition and tear film mixing in spontaneous blinking. Contact Lens Anterior Eye (2015), http://dx.doi.org/10.1016/j.clae.2015.01.012
