1040-5488/14/9112-1406/0 VOL. 91, NO. 12, PP. 1406Y1411 OPTOMETRY AND VISION SCIENCE Copyright * 2014 American Academy of Optometry
ORIGINAL ARTICLE
Tear Film Dynamics on Soft Contact Lenses Dorota H. Szczesna-Iskander* and D. Robert Iskander†
ABSTRACT Purpose. To analyze and model the dynamics of tear film surface quality (TFSQ) in a group of subjects with healthy eyes, before and during contact lens (CL) wear, and in a group of subjects with dry eyes (DEs). Methods. Four sets of TFSQ measurements with lateral shearing interferometry were retrospectively analyzed on two groups of subjects. The first group included 13 CL wearers for which TFSQ measurements corresponding to baseline, Best CL, and Worst CL were selected. The second group included 13 DE subjects. The acquired TFSQ time series were fit with a powerexponential model. Tear film surface quality stability time, best TFSQ, and integrated poststability time characterizing the tear film deterioration process were derived. Results. The proposed power-exponential model was better suited (higher correlation values) for the TFSQ dynamics on CL rather than for those of baseline and DE measurements. The average baseline TFSQ Stability Time was significantly longer (p G 0.001) than those in the DE group and with both the best and worse CL. The average Best TFSQ achieved at baseline was statistically significantly better (p = 0.03) than that for the DE group. The average best TFSQ was significantly better (p G 0.01) for the Best CL than for that of the Worst CL. Deterioration of TFSQ on both best and Worst CL was substantially faster than that achieved for the DE group. Conclusions. The observed distinct change in the sign of the TFSQ velocity on contact lenses suggests a two-phase dynamics in which the postblink stability phase is followed by a phase of dewetting. Lens material properties influence the first phase but play little role after the dewetting process occurs. (Optom Vis Sci 2014;91:1406Y1411) Key Words: tear film dynamics, tear film thinning, tear film break-up, dewetting, contact lens material
T
he tear film dynamics on the eye are well understood.1 Cycles of turnover, evaporation, and absorption,2,3 specific phases of tear film distribution on the eye,4 impact on ocular aberrations,5,6 and factors such as sex and age7,8 have been extensively studied and modeled9Y11 in both healthy and dry eyes (DEs).12,13 The dynamics of the tear film on soft contact lenses (CLs) has been studied in terms of tear film distribution and thinning rates but generally is less explored than that of precorneal tear film. Bruce et al.14 studied the break-up locations and noted that, unlike the precorneal tear film, the distribution of break-ups on contact lenses was less regular. Nichols et al.15 studied thinning rates of the precorneal and prelens tear films using interferometry and concluded that the higher rates of prelens tear film thinning might be attributed to dewetting rather than to evaporation. Liu et al.16 associated these break-ups with symptoms of blurry vision in CL wearers.
*PhD † PhD, DSc Institute of Physics, Wroclaw University of Technology, Wroclaw, Poland (DHS-I); and Institute of Biomedical Engineering and Instrumentation, Wroclaw University of Technology, Wroclaw, Poland (DRI).
The tear film thinning process is thought to be guided by three main mechanisms,3,15,17 including an inward flow of water into the epithelium or CL, tangential flow caused by surface tension gradients, and evaporation. It has been suggested that the inward flow plays a lesser role than the other two mechanisms and that evaporation seems to be the strongest factor influencing the tear film thinning process.15 The increased evaporation and thinning rate are observed in DEs as well as on CLs. However, in DEs, this phenomenon is mostly related to the composition of tear film; whereas in CL wearers, the interaction between tear film and lens material might play an additional role.18 The presence of a soft CL on eye alters the properties of tear film and hence has an impact on tear film dynamics.19Y22 The effect of CL wear is observed at both a biochemical as well as biophysical level, resulting in changes in structural stability of the tear film. Of particular interest is the process of CL surface dewetting23 that is related to CL wettability but it essentially describes the rupture of a thin prelens tear film.24 Precorneal and prelens dynamics can be noninvasively observed with lateral shearing interferometry,4,13,18 which provides reliable time-varying estimates of tear film surface quality (TFSQ). Following Nichols et al.,15 it was of interest to ascertain whether the
Optometry and Vision Science, Vol. 91, No. 12, December 2014
Copyright © American Academy of Optometry. Unauthorized reproduction of this article is prohibited.
Tear Film Dynamics on Soft Contact LensesVSzczesna-Iskander and Iskander
observed dynamics in TFSQ in CL wearers corresponded to the processes of evaporation or dewetting. Hence, the aim was to analyze TFSQ measured in suppressed blinking conditions and to describe its dynamics in a group of normal healthy eyes, before and during soft CL wear, and in DEs. Of particular interest was to find an appropriate model that would describe these changes.
MATERIALS AND METHODS Subjects The authors retrospectively analyzed the TFSQ data of two groups of subjects: the CL group consisting of 13 subjects aged 25 to 47 years (mean T SD age, 32 T 7 years) who wore daily disposable CLs and the DE group consisting of 13 subjects aged 20 to 68 years (mean T SD age, 47 T 19 years) who were diagnosed as having DE syndrome or assessed as having a marginal DE. These two groups of subjects participated in two separate experiments. All participants reviewed and signed informed consent documents and were treated according to the Declaration of Helsinki. A standard clinical assessment of DE symptoms and signs was performed on all subjects in both groups. The particulars of these examinations were previously described in detail.13,18 In summary, these tests included a clinical history, McMonnies questionnaire,25 slit lamp biomicroscopy examination, phenol red thread test of tear volume, fluorescein tear film break-up time test (FTBUT), and assessment of ocular surface staining with fluorescein and lissamine green dyes grading according to the National Eye Institute grading scales. Dry eye was diagnosed if the subject exhibited two characteristics of global DE definition {McMonnies Questionnaire scores 914, FBUT score G10 s as an average from three measurements, and/or significant ocular surface staining with fluorescein and lissamine green [staining score 93 (out of 12)]}. According to this classification, in the DE group, there were 11 subjects (10 women and 1 man) diagnosed as having DE (six subjects had lipid anomaly, four aqueous deficiency, and one atopic DE) and two women as having marginal DE (based on FBUT G10 s and ocular surface staining). In the group of subjects who wore CLs (three women, 10 men), none had any ocular disease, 12 subjects were assessed as having a normal tear film and ocular surface, and 1 subject exhibited a reduced FTBUT (mean FTBUT, 7.2 s). His signs of corneal surface staining (National Eye Institute score) were equal to three. One subject in the group was a regular soft CL wearer on a daily basis. All subjects were asked not to wear lenses for a week before the commencement of the experiment and not to use any eye drops on the day of the experiment.
Measurement Technique Two techniques were used for noninvasive TFSQ assessment: High-Speed Videokeratoscopy and Lateral Shearing Interferometry (LSI).13,18 Because, in both experiments, LSI methodology indicated higher sensitivity to subtle changes in TFSQ, we focused here on the analysis of images recorded using the LSI technique. The principle of the measurement is based on the numerical analysis of wavefront reflected from tear film surface covering an eye or a CL. Central tear film surface area of about 4 mm in diameter is imaged. The superposition of two waves produced by a shearing
1407
element in the instrument results in a wave pattern called an interferogram. The interference fringes are regular and even when the wavefront is reflected from a smooth regular tear film surface. Any irregularity on the surface deforms the reflected wave that converts into some disturbance in the interferogram. The CCD camera records 25 frames per second. Every frame (~7 Km of lateral resolution) shows a pattern of interference fringes corresponding to the surface of the tear film at that given moment. For an objective assessment of TFSQ, a measure has been proposed based on analysis of the frequency spectrum of each interferogram. Each frame can be described by an arbitrary number, referred to as the TFSQ parameter, which is inversely proportional to the interferometric pattern disturbance. The details of the measurement and analysis techniques have been explained in our previous work.4,26
Experiment In both experiments, the data were collected in natural and suppressed blinking conditions. Changes in TFSQ observed in suppressed blinking conditions were of interest in this article. At the beginning of each of the 20-s measurements, the subject was asked to blink and then focus on the instrument’s fixation target and keep his or her eyes open as long as possible. In the DE group, three video sequences between 3:00 p.m. and 4:00 p.m. were collected with about 3-minute break between each measurement. In the CL group, two in vivo TFSQ measurements were taken in the morning (between 8:00 a.m. and 10:00 a.m.) and in the afternoon (between 4:00 p.m. and 6:00 p.m.) first on the cornea and then on a separate day on the CL. In the experiment with the CL group,18 each subject was wearing a pair of four different daily disposable CLs of power j0.5D on nonconsecutive days. The TFSQ assessments were performed on the bare eye, 30 minutes after insertion in the morning and after 8 hours of lens wear. Based on the objective numerical analysis of TFSQ assessed in natural blinking conditions, two lens materials were chosen individually for each subject that caused the least and the greatest decline in the average TFSQ with respect to that of the precorneal tear film. Only the results obtained for those two lenses (i.e., the best and the worst for a given subject) were considered in this article.
Model of Tear Film Surface Dynamics Previously, we proposed a multilinear model of precorneal TFSQ dynamics assessed in LSI-based measurements on DEs as well as on eyes with a normal tear film.4 The aim was to apply a parsimonious model that could determine phases of tear film dynamics on the bare eye in suppressed blinking conditions and define their transition points. The slope of linear function was used to describe each phase of tear film dynamics. Using the trilinear and quadlinear models, which were statistically compared, the two-stage tear film buildup process was distinguished that was followed by the tear film stability phase and/or the tear film thinning phase. In normal eyes, the stability phase was sometimes observed until the end of the recorded sequence or the next blink, and the thinning phase was in most cases indistinguishable. In DEs, on the other hand, the thinning phase followed either directly the buildup process or its dynamics was more pronounced than that of normal eyes.
Optometry and Vision Science, Vol. 91, No. 12, December 2014
Copyright © American Academy of Optometry. Unauthorized reproduction of this article is prohibited.
1408 Tear Film Dynamics on Soft Contact LensesVSzczesna-Iskander and Iskander
A pattern of tear film dynamics observed on CLs was sometimes different to that of bare eyes. The differences were observed in the buildup, stability, and the thinning phases of tear film dynamics. The buildup and stability phases were much shorter on CLs than on bare eyes and sometimes they were indistinguishable, forming one phase of tear film dynamics (see the model of TFSQ dynamics in Fig. 1). The tear film thinning phase immediately followed the previous smoothing phase, and it was often characterized by a rapid increase of TFSQ that was later saturating to a level, likely corresponding to a state where a very thin layer of tear was covering the surface of the lens. Hence, previous suggested multilinear models were inadequate to describe the abrupt change in TFSQ estimator on CLs. Therefore, a combination of exponential and power models was proposed. The power function ( f 1 ¼ a1 t b1 þ c1 ) was used to define the first two phases of tear film dynamics: buildup and the stability phase (Fig. 1). The exponential function ( f 2 ¼ a2 eb2 t þ c2 ) was used to describe the tear film thinning phase and that of progressive tear film deterioration. The aim of applying this model was to extract parameters that specify the observed tear film dynamics. The two considered functions were fitted to the raw TFSQ data using an iterative constrained least-squares algorithm similar to the bilinear and multilinear models described earlier,4 where the exponential function f2 was conditioned on the estimator of the power function F1. The interception point of two functions denotes the Best TFSQ (marked on the y axis in Fig. 1) that was achieved in the recording. It marks the end of the combined buildup and stability phases and corresponds to the TFSQ Stability Time (marked on the x axis in Fig. 1). To compare the early changes of the exponential function f2 between the groups of subjects, the area under the curve (AUC) was calculated numerically using trapezoidal
integration for the first 4 s after the intersection point using a step of 1 s. The area was normalized to Best TFSQ value (Fig. 1).
Statistical Analysis Because several measurements were collected for each subject, the results were averaged first for the subject, and then the group mean and median were calculated to see whether there was a change in the central tendency of the data. There was not enough data to reliably test for normality. Pairwise comparison between the groups was performed using Wilcoxon signed rank test. A value of p G 0.05 was considered statistically significant. To assess the goodness of fit of the model, the estimated root mean square (RMS) value and the Pearson correlation coefficient were evaluated.
RESULTS Several representative examples are shown in Fig. 2, illustrating changes of TFSQ estimator analyzed in suppressed blinking conditions. The TFSQ dynamics on bare eyes are depicted in Fig. 2A, whereas those of CL material assessed as the best (Best CL) for the individual subject and contact lens material that induced the highest impairment in TFSQ (Worst CL) are shown in Fig. 2B and C, respectively. For most of the DE subjects, the time of 20 s was too long to keep their eyes open, so the recordings in this group were usually shorter. The TFSQ dynamics for a subject with DE, with lipid anomaly diagnosed based on meibomian gland dysfunction grade 2.5, is shown in Fig. 2D. Notice that the scales in the y axes are different for each of the plots. Data of the CL group were distributed into three subcategories: Bare Eye, Best CL, and Worst CL. The parameters extracted from the proposed model to characterize the average TFSQ behavior
FIGURE 1. The model of tear film surface dynamics on contact lenses. The first phase (ending at TFSQ Stability Time) is modeled with a power function, whereas the second phase is modeled by an exponential function. Optometry and Vision Science, Vol. 91, No. 12, December 2014
Copyright © American Academy of Optometry. Unauthorized reproduction of this article is prohibited.
Tear Film Dynamics on Soft Contact LensesVSzczesna-Iskander and Iskander
1409
as well as to assess the goodness of fit of the power-exponential model in the four considered subcategories of subjects are summarized in Table 1. The average TFSQ Stability Time in the Bare Eye subcategory was 11.95 T 5.62 s (mean T SD), which was significantly longer (p G 0.001) than in the DE group and both Best and Worst CL subcategories (3.95 T 2.87 s, 4.28 T 2.25 s, and 2.94 T 1.14 s, respectively). No statistically significant differences were found between the DE group, Best CL, and the Worst CL subcategories in the average TFSQ Stability Time estimator. Fig. 3 shows the results of the average Best TFSQ for the four considered subcategories. The lowest values of the TFSQ estimator was achieved for the Bare Eye subcategory (105.68 T 6.03), and the average TFSQ was statistically significantly better (p = 0.03) than that of the DE group (114.80 T 12.44). Next, the average Best TFSQ was significantly better (p G 0.01) for the Best CL subcategory (118.10 T 7.94) than for that of the Worst CL (141.95 T 18.45). There was no statistically significant difference in the average Best TFSQ between the DE group and the Best CL subcategory. The group average RMS value was the highest for the Worst CL subcategory (9.65 T 0.90) and the lowest for the Bare Eye subcategory (7.37 T 1.80). Similarly, the estimated Pearson correlation coefficient was the highest (0.85 T 0.14) and the lowest (0.52 T 0.27) for the Worst CL and Bare Eye subcategories, respectively. The latter indicates that the proposed model is better suited for the CL tear film dynamics than for the bare eye scenarios. The former simply means that the variability of TFSQ in Bare Eye measurements is lower than those in CL acquisitions. The differences in the early changes in the poststability phase within the first 4 s are presented in Fig. 4. The AUC increases similarly for both the Best CL and the Worst CL subcategories. The slope of a linear approximation to AUC is 7.82 and 6.91 for Best CL and Worst CL subcategories, respectively. A substantially smaller change was achieved for the DE group with a linear slope of 4.96. This indicates that the tear film behaves better on the physiologic surface of the DE than on the artificial surface of a CL. Note that, for most of the normal eyes, the first power function was fitted to almost the complete sequence of data as indicated in Fig. 2A; therefore, this group was not taken into consideration in the analysis of AUC.
DISCUSSION
FIGURE 2. Examples of fitting the iterative power-exponential model to data recorded on bare eye with normal tear film (A), the best contact lens material (B), the worst contact lens material (C), and a dry eye (D).
Lateral shearing interferometry has the capability to observe subtle changes in the tear film surface as evidenced in our earlier studies. This technique of measurement performed well in distinguishing DEs from those having a normal tear film,13 and it showed the ability to differentiate CL materials through the analysis of the average TFSQ in natural blinking conditions.18 When the TFSQ is measured in suppressed blinking conditions on a CL, its dynamics essentially follow a two-phase process in contrast to bare eye or DE measurements that conform, in most cases, to the previously suggested multilinear model consisting of up to five distinct phases.4 A new model has been proposed to describe the two-phase process of tear film dynamics on contact lenses. The first phase of this process is well modeled by a power function, whereas the second phase is better fitted by an exponential function that is
Optometry and Vision Science, Vol. 91, No. 12, December 2014
Copyright © American Academy of Optometry. Unauthorized reproduction of this article is prohibited.
1410 Tear Film Dynamics on Soft Contact LensesVSzczesna-Iskander and Iskander TABLE 1.
Comparison of four parameters received from the proposed model of TFSQ behavior analyzed in four conditions: on the Best CL, Worst CL, DE, and normal eye Parameter
Eye
Best CL
Worst CL
DE
2.94 T 1.14 2.99
3.95 T 2.87 4.16
141.95 T 18.45 148.57
114.80 T 12.44 110.08
TFSQ stability time
Mean T SD Median
11.95 T 5.63 11.43
4.28 T 2.25 3.65
Best TFSQ
Mean T SD Median
105.68 T 6.03 104.90
118.10 T 7.94 120.73
Min RMS
Mean T SD Median
7.37 T 1.80 6.19
8.76 T 1.31 8.99
9.65 T 0.90 9.68
9.63 T 1.88 9.35
R2
Mean T SD Median
0.52 T 0.27 0.56
0.80 T 0.17 0.90
0.85 T 0.14 0.90
0.75 T 0.18 0.77
TFSQ, Tear Film Surface Quality; RMS, root mean square; R 2, coefficient of determination; Best CL, the best contact lens material; Worst CL, the worst contact lens material; DE, dry eye; Eye, bare eye.
conditioned on the first power function. The two types of functions were chosen empirically and found to be better suited to the tear film dynamics on CLs than to those of the bare eye and DE. In the majority of the recorded interferogram sequences of prelens TFSQ, a stage is observed, in which the estimated TFSQ abruptly changes. This suggests that, after a blink, the tear film on CL quickly smooths out to achieve the best possible (materialdependent) TFSQ and then follows a phase of sudden deterioration. We attribute the phenomenon of a sudden deterioration phase of the tear film on the CL to the wetting characteristics of the lens27 and particularly to the process of dewetting.15 Note that, in some of the measurements, the tear film deterioration phase caused the disappearance of interferogram fringe information, suggesting complete dryness of the lens surface. Such a phase is not normally observed in the case of bare eye measurements (both normal eyes and DEs) where the tear film thinning phase is much more subtle and more defined. It is suggested that the precorneal behavior of the tear film thinning is mainly caused by the evaporation but also related to the tangential flow surface tension gradients.3,15,17 Based on the previously described experiment,18 where four daily disposable CLs were tested, the best and worst lens
measurements acquired in natural blinking conditions (that caused the least and highest declines in TFSQ, respectively) were chosen for each of the subjects. The differences in tear film quality between those lens materials were confirmed by the results of group statistics of parameters extracted from the proposed model. The average Best TFSQ was significantly better for the Best CL than for the Worst CL. The time of TFSQ stability was also longer for the Best CL; however, the difference was not statistically significant. Shorter stability time of the tear film on contact lenses, rather than on the bare eye, confirmed earlier observations.28,29 The particular type of the lens material did play some role in the first phase of TFSQ dynamics affecting both the TFSQ Stability Time and the Best TFSQ. However, regardless of the initial interaction between lens material and tear film, the study showed that, when destabilization of the tear film surface starts, it progresses similarly quickly on the Best and the Worst CLs. This suggests that the material properties of a CL have little influence on tear film dynamics after the process of dewetting occurs. It is worth noting that the transition point from the first to the second phase of TFSQ dynamics was on average less than 3 s for the Worst CL and just above 4 s for the Best CLVvalues similar to those achieved for the DE group and shorter than an average interblink interval.30 Developing contact lens materials that would delay this
FIGURE 3.
FIGURE 4.
The best TFSQ_Av in four groups of ocular conditions. Statistically significant differences are in bold font and marked with asterisks. Error bars indicate 1 SD.
The dynamics of TFSQ in the first 4 s after the intersection point illustrates the progressive tear film thinning process.
Optometry and Vision Science, Vol. 91, No. 12, December 2014
Copyright © American Academy of Optometry. Unauthorized reproduction of this article is prohibited.
Tear Film Dynamics on Soft Contact LensesVSzczesna-Iskander and Iskander
transition point could result in the future in a better quality of the prelens tear film.
ACKNOWLEDGMENTS Supported by the European Regional Development Fund within Innovative Economy Operational Program cofinanced by the Foundation for Polish Science (POMOST/2012-5/8/0072 to DS-I). Neither author has anything to disclose. Received May 2, 2014; accepted July 29, 2014.
REFERENCES 1. Tomlinson A, Khanal S. Assessment of tear film dynamics: quantification approach. Ocul Surf 2005;3:81Y95. 2. Tsubota K. Tear dynamics and dry eye. Prog Retin Eye Res 1998; 17:565Y96. 3. King-Smith PE, Nichols JJ, Nichols KK, Fink BA, Braun RJ. Contributions of evaporation and other mechanisms to tear film thinning and break-up. Optom Vis Sci 2008;85:623Y30. 4. Szczesna DH, Iskander DR. Lateral shearing interferometry for analysis of tear film surface kinetics. Optom Vis Sci 2010;87:513Y7. 5. Li KY, Yoon G. Changes in aberrations and retinal image quality due to tear film dynamics. Opt Express 2006;14:12552Y9. 6. Koh S, Maeda N. Wavefront sensing and the dynamics of tear film. Cornea 2007;26:S41Y5. 7. Maissa C, Guillon M. Tear film dynamics and lipid layer characteristics– effect of age and gender. Cont Lens Anterior Eye 2010;33:176Y82. 8. Ozdemir M, Temizdemir H. Age- and gender-related tear function changes in normal population. Eye (Lond) 2010;24:79Y83. 9. Jones MB, Please CP, McElwain DL, Fulford GR, Roberts AP, Collins MJ. Dynamics of tear film deposition and draining. Math Med Biol 2005;22:265Y88. 10. Braun RJ. Dynamics of the tear film. Annu Rev Fluid Mech 2012; 44:267Y97. 11. Braun RJ, Gewecke NR, Begley CG, King-Smith PE, Siddique JI. A model for tear film thinning with osmolarity and fluorescein. Invest Ophthalmol Vis Sci 2014;55:1133Y42. 12. Monte´s-Mico´ R1, Alio´ JL, Charman WN. Dynamic changes in the tear film in dry eyes. Invest Ophthalmol Vis Sci 2005;46:1615Y9. 13. Szczesna DH, Alonso-Caneiro D, Iskander DR, Read SA, Collins MJ. Predicting dry eye using noninvasive techniques of tear film surface assessment. Invest Ophthalmol Vis Sci 2011;52:751Y6. 14. Bruce AS, Mainstone JC, Golding TR. Analysis of tear film breakup on Etafilcon A hydrogel lenses. Biomaterials 2001;22:3249Y56. 15. Nichols JJ, Mitchell GL, King-Smith PE. Thinning rate of the precorneal and prelens tear films. Invest Ophthalmol Vis Sci 2005;46:2353Y61.
1411
16. Liu H, Thibos L, Begley CG, Bradley A. Measurement of the time course of optical quality and visual deterioration during tear break-up. Invest Ophthalmol Vis Sci 2010;51:3318Y26. 17. King-Smith PE, Hinel EA, Nichols JJ. Application of a novel interferometric method to investigate the relation between lipid layer thickness and tear film thinning. Invest Ophthalmol Vis Sci 2010; 51:2418Y23. 18. Szczesna-Iskander DH, Iskander DR, Read SA, Alonso-Caneiro D. Noninvasive in vivo assessment of soft contact lens type on tear film surface quality. Invest Ophthalmol Vis Sci 2012;53:525Y31. 19. Guillon JP. Tear film structure and contact lens. In: Holly FJ, Lamberts DW, MacKeen DL, eds. The Preocular Tear Film in Health, Disease, and Contact Lens Wear. Lubbock, TX: Dry Eye Institute; 1986:914Y39. 20. Guillon M, Maissa C, Girard-Claudon K, Cooper P. Influence of the tear film composition on tear film structure and symptomatology of soft contact lens wearers. Adv Exp Med Biol 2002;506:895Y9. 21. Korb DR. Tear film-contact lens interactions. Adv Exp Med Biol 1994;350:403Y10. 22. Rohit A, Willcox M, Stapleton F. Tear lipid layer and contact lens comfort: a review. Eye Contact Lens 2013;39:247Y53. 23. Varikooty J, Keir N, Simpson T. Estimating tear film spread and stability through tear hydrodynamics. Optom Vis Sci 2012;89: E1119Y24. 24. Karapanagiotis I, Gerberich WW. Polymer film rupturing in comparison with leveling and dewetting. Surf Sci 2005;594:192Y202. 25. McMonnies CW. Key questions in a dry eye history. J Am Optom Assoc 1986;57:512Y7. 26. Szczesna DH, Iskander DR. Robust estimation of tear film surface quality in lateral shearing interferometry. J Biomed Opt 2009; 14:064039. 27. Fagehi R, Tomlinson A, Manahilov V, Haddad M. Contact lens in vitro wettability by interferometry measures of drying dynamics. Eye Contact Lens 2013;39:365Y75. 28. Fonn D, Situ P, Simpson T. Hydrogel lens dehydration and subjective comfort and dryness ratings in symptomatic and asymptomatic contact lens wearers. Optom Vis Sci 1999;76:700Y4. 29. Guillon M, Maissa C. Contact lens wear affects tear film evaporation. Eye Contact Lens 2008;34:326Y30. 30. Johnston PR, Rodriguez J, Lane KJ, Ousler G, Abelson MB. The interblink interval in normal and dry eye subjects. Clin Ophthalmol 2013;7:253Y9.
Dorota Szczesna-Iskander Wroclaw University of Technology Wybrzeze Wyspianskiego 27 50-370 Wroclaw Poland e-mail: [email protected]
Optometry and Vision Science, Vol. 91, No. 12, December 2014
Copyright © American Academy of Optometry. Unauthorized reproduction of this article is prohibited.
ORIGINAL ARTICLE
Tear Film Dynamics on Soft Contact Lenses Dorota H. Szczesna-Iskander* and D. Robert Iskander†
ABSTRACT Purpose. To analyze and model the dynamics of tear film surface quality (TFSQ) in a group of subjects with healthy eyes, before and during contact lens (CL) wear, and in a group of subjects with dry eyes (DEs). Methods. Four sets of TFSQ measurements with lateral shearing interferometry were retrospectively analyzed on two groups of subjects. The first group included 13 CL wearers for which TFSQ measurements corresponding to baseline, Best CL, and Worst CL were selected. The second group included 13 DE subjects. The acquired TFSQ time series were fit with a powerexponential model. Tear film surface quality stability time, best TFSQ, and integrated poststability time characterizing the tear film deterioration process were derived. Results. The proposed power-exponential model was better suited (higher correlation values) for the TFSQ dynamics on CL rather than for those of baseline and DE measurements. The average baseline TFSQ Stability Time was significantly longer (p G 0.001) than those in the DE group and with both the best and worse CL. The average Best TFSQ achieved at baseline was statistically significantly better (p = 0.03) than that for the DE group. The average best TFSQ was significantly better (p G 0.01) for the Best CL than for that of the Worst CL. Deterioration of TFSQ on both best and Worst CL was substantially faster than that achieved for the DE group. Conclusions. The observed distinct change in the sign of the TFSQ velocity on contact lenses suggests a two-phase dynamics in which the postblink stability phase is followed by a phase of dewetting. Lens material properties influence the first phase but play little role after the dewetting process occurs. (Optom Vis Sci 2014;91:1406Y1411) Key Words: tear film dynamics, tear film thinning, tear film break-up, dewetting, contact lens material
T
he tear film dynamics on the eye are well understood.1 Cycles of turnover, evaporation, and absorption,2,3 specific phases of tear film distribution on the eye,4 impact on ocular aberrations,5,6 and factors such as sex and age7,8 have been extensively studied and modeled9Y11 in both healthy and dry eyes (DEs).12,13 The dynamics of the tear film on soft contact lenses (CLs) has been studied in terms of tear film distribution and thinning rates but generally is less explored than that of precorneal tear film. Bruce et al.14 studied the break-up locations and noted that, unlike the precorneal tear film, the distribution of break-ups on contact lenses was less regular. Nichols et al.15 studied thinning rates of the precorneal and prelens tear films using interferometry and concluded that the higher rates of prelens tear film thinning might be attributed to dewetting rather than to evaporation. Liu et al.16 associated these break-ups with symptoms of blurry vision in CL wearers.
*PhD † PhD, DSc Institute of Physics, Wroclaw University of Technology, Wroclaw, Poland (DHS-I); and Institute of Biomedical Engineering and Instrumentation, Wroclaw University of Technology, Wroclaw, Poland (DRI).
The tear film thinning process is thought to be guided by three main mechanisms,3,15,17 including an inward flow of water into the epithelium or CL, tangential flow caused by surface tension gradients, and evaporation. It has been suggested that the inward flow plays a lesser role than the other two mechanisms and that evaporation seems to be the strongest factor influencing the tear film thinning process.15 The increased evaporation and thinning rate are observed in DEs as well as on CLs. However, in DEs, this phenomenon is mostly related to the composition of tear film; whereas in CL wearers, the interaction between tear film and lens material might play an additional role.18 The presence of a soft CL on eye alters the properties of tear film and hence has an impact on tear film dynamics.19Y22 The effect of CL wear is observed at both a biochemical as well as biophysical level, resulting in changes in structural stability of the tear film. Of particular interest is the process of CL surface dewetting23 that is related to CL wettability but it essentially describes the rupture of a thin prelens tear film.24 Precorneal and prelens dynamics can be noninvasively observed with lateral shearing interferometry,4,13,18 which provides reliable time-varying estimates of tear film surface quality (TFSQ). Following Nichols et al.,15 it was of interest to ascertain whether the
Optometry and Vision Science, Vol. 91, No. 12, December 2014
Copyright © American Academy of Optometry. Unauthorized reproduction of this article is prohibited.
Tear Film Dynamics on Soft Contact LensesVSzczesna-Iskander and Iskander
observed dynamics in TFSQ in CL wearers corresponded to the processes of evaporation or dewetting. Hence, the aim was to analyze TFSQ measured in suppressed blinking conditions and to describe its dynamics in a group of normal healthy eyes, before and during soft CL wear, and in DEs. Of particular interest was to find an appropriate model that would describe these changes.
MATERIALS AND METHODS Subjects The authors retrospectively analyzed the TFSQ data of two groups of subjects: the CL group consisting of 13 subjects aged 25 to 47 years (mean T SD age, 32 T 7 years) who wore daily disposable CLs and the DE group consisting of 13 subjects aged 20 to 68 years (mean T SD age, 47 T 19 years) who were diagnosed as having DE syndrome or assessed as having a marginal DE. These two groups of subjects participated in two separate experiments. All participants reviewed and signed informed consent documents and were treated according to the Declaration of Helsinki. A standard clinical assessment of DE symptoms and signs was performed on all subjects in both groups. The particulars of these examinations were previously described in detail.13,18 In summary, these tests included a clinical history, McMonnies questionnaire,25 slit lamp biomicroscopy examination, phenol red thread test of tear volume, fluorescein tear film break-up time test (FTBUT), and assessment of ocular surface staining with fluorescein and lissamine green dyes grading according to the National Eye Institute grading scales. Dry eye was diagnosed if the subject exhibited two characteristics of global DE definition {McMonnies Questionnaire scores 914, FBUT score G10 s as an average from three measurements, and/or significant ocular surface staining with fluorescein and lissamine green [staining score 93 (out of 12)]}. According to this classification, in the DE group, there were 11 subjects (10 women and 1 man) diagnosed as having DE (six subjects had lipid anomaly, four aqueous deficiency, and one atopic DE) and two women as having marginal DE (based on FBUT G10 s and ocular surface staining). In the group of subjects who wore CLs (three women, 10 men), none had any ocular disease, 12 subjects were assessed as having a normal tear film and ocular surface, and 1 subject exhibited a reduced FTBUT (mean FTBUT, 7.2 s). His signs of corneal surface staining (National Eye Institute score) were equal to three. One subject in the group was a regular soft CL wearer on a daily basis. All subjects were asked not to wear lenses for a week before the commencement of the experiment and not to use any eye drops on the day of the experiment.
Measurement Technique Two techniques were used for noninvasive TFSQ assessment: High-Speed Videokeratoscopy and Lateral Shearing Interferometry (LSI).13,18 Because, in both experiments, LSI methodology indicated higher sensitivity to subtle changes in TFSQ, we focused here on the analysis of images recorded using the LSI technique. The principle of the measurement is based on the numerical analysis of wavefront reflected from tear film surface covering an eye or a CL. Central tear film surface area of about 4 mm in diameter is imaged. The superposition of two waves produced by a shearing
1407
element in the instrument results in a wave pattern called an interferogram. The interference fringes are regular and even when the wavefront is reflected from a smooth regular tear film surface. Any irregularity on the surface deforms the reflected wave that converts into some disturbance in the interferogram. The CCD camera records 25 frames per second. Every frame (~7 Km of lateral resolution) shows a pattern of interference fringes corresponding to the surface of the tear film at that given moment. For an objective assessment of TFSQ, a measure has been proposed based on analysis of the frequency spectrum of each interferogram. Each frame can be described by an arbitrary number, referred to as the TFSQ parameter, which is inversely proportional to the interferometric pattern disturbance. The details of the measurement and analysis techniques have been explained in our previous work.4,26
Experiment In both experiments, the data were collected in natural and suppressed blinking conditions. Changes in TFSQ observed in suppressed blinking conditions were of interest in this article. At the beginning of each of the 20-s measurements, the subject was asked to blink and then focus on the instrument’s fixation target and keep his or her eyes open as long as possible. In the DE group, three video sequences between 3:00 p.m. and 4:00 p.m. were collected with about 3-minute break between each measurement. In the CL group, two in vivo TFSQ measurements were taken in the morning (between 8:00 a.m. and 10:00 a.m.) and in the afternoon (between 4:00 p.m. and 6:00 p.m.) first on the cornea and then on a separate day on the CL. In the experiment with the CL group,18 each subject was wearing a pair of four different daily disposable CLs of power j0.5D on nonconsecutive days. The TFSQ assessments were performed on the bare eye, 30 minutes after insertion in the morning and after 8 hours of lens wear. Based on the objective numerical analysis of TFSQ assessed in natural blinking conditions, two lens materials were chosen individually for each subject that caused the least and the greatest decline in the average TFSQ with respect to that of the precorneal tear film. Only the results obtained for those two lenses (i.e., the best and the worst for a given subject) were considered in this article.
Model of Tear Film Surface Dynamics Previously, we proposed a multilinear model of precorneal TFSQ dynamics assessed in LSI-based measurements on DEs as well as on eyes with a normal tear film.4 The aim was to apply a parsimonious model that could determine phases of tear film dynamics on the bare eye in suppressed blinking conditions and define their transition points. The slope of linear function was used to describe each phase of tear film dynamics. Using the trilinear and quadlinear models, which were statistically compared, the two-stage tear film buildup process was distinguished that was followed by the tear film stability phase and/or the tear film thinning phase. In normal eyes, the stability phase was sometimes observed until the end of the recorded sequence or the next blink, and the thinning phase was in most cases indistinguishable. In DEs, on the other hand, the thinning phase followed either directly the buildup process or its dynamics was more pronounced than that of normal eyes.
Optometry and Vision Science, Vol. 91, No. 12, December 2014
Copyright © American Academy of Optometry. Unauthorized reproduction of this article is prohibited.
1408 Tear Film Dynamics on Soft Contact LensesVSzczesna-Iskander and Iskander
A pattern of tear film dynamics observed on CLs was sometimes different to that of bare eyes. The differences were observed in the buildup, stability, and the thinning phases of tear film dynamics. The buildup and stability phases were much shorter on CLs than on bare eyes and sometimes they were indistinguishable, forming one phase of tear film dynamics (see the model of TFSQ dynamics in Fig. 1). The tear film thinning phase immediately followed the previous smoothing phase, and it was often characterized by a rapid increase of TFSQ that was later saturating to a level, likely corresponding to a state where a very thin layer of tear was covering the surface of the lens. Hence, previous suggested multilinear models were inadequate to describe the abrupt change in TFSQ estimator on CLs. Therefore, a combination of exponential and power models was proposed. The power function ( f 1 ¼ a1 t b1 þ c1 ) was used to define the first two phases of tear film dynamics: buildup and the stability phase (Fig. 1). The exponential function ( f 2 ¼ a2 eb2 t þ c2 ) was used to describe the tear film thinning phase and that of progressive tear film deterioration. The aim of applying this model was to extract parameters that specify the observed tear film dynamics. The two considered functions were fitted to the raw TFSQ data using an iterative constrained least-squares algorithm similar to the bilinear and multilinear models described earlier,4 where the exponential function f2 was conditioned on the estimator of the power function F1. The interception point of two functions denotes the Best TFSQ (marked on the y axis in Fig. 1) that was achieved in the recording. It marks the end of the combined buildup and stability phases and corresponds to the TFSQ Stability Time (marked on the x axis in Fig. 1). To compare the early changes of the exponential function f2 between the groups of subjects, the area under the curve (AUC) was calculated numerically using trapezoidal
integration for the first 4 s after the intersection point using a step of 1 s. The area was normalized to Best TFSQ value (Fig. 1).
Statistical Analysis Because several measurements were collected for each subject, the results were averaged first for the subject, and then the group mean and median were calculated to see whether there was a change in the central tendency of the data. There was not enough data to reliably test for normality. Pairwise comparison between the groups was performed using Wilcoxon signed rank test. A value of p G 0.05 was considered statistically significant. To assess the goodness of fit of the model, the estimated root mean square (RMS) value and the Pearson correlation coefficient were evaluated.
RESULTS Several representative examples are shown in Fig. 2, illustrating changes of TFSQ estimator analyzed in suppressed blinking conditions. The TFSQ dynamics on bare eyes are depicted in Fig. 2A, whereas those of CL material assessed as the best (Best CL) for the individual subject and contact lens material that induced the highest impairment in TFSQ (Worst CL) are shown in Fig. 2B and C, respectively. For most of the DE subjects, the time of 20 s was too long to keep their eyes open, so the recordings in this group were usually shorter. The TFSQ dynamics for a subject with DE, with lipid anomaly diagnosed based on meibomian gland dysfunction grade 2.5, is shown in Fig. 2D. Notice that the scales in the y axes are different for each of the plots. Data of the CL group were distributed into three subcategories: Bare Eye, Best CL, and Worst CL. The parameters extracted from the proposed model to characterize the average TFSQ behavior
FIGURE 1. The model of tear film surface dynamics on contact lenses. The first phase (ending at TFSQ Stability Time) is modeled with a power function, whereas the second phase is modeled by an exponential function. Optometry and Vision Science, Vol. 91, No. 12, December 2014
Copyright © American Academy of Optometry. Unauthorized reproduction of this article is prohibited.
Tear Film Dynamics on Soft Contact LensesVSzczesna-Iskander and Iskander
1409
as well as to assess the goodness of fit of the power-exponential model in the four considered subcategories of subjects are summarized in Table 1. The average TFSQ Stability Time in the Bare Eye subcategory was 11.95 T 5.62 s (mean T SD), which was significantly longer (p G 0.001) than in the DE group and both Best and Worst CL subcategories (3.95 T 2.87 s, 4.28 T 2.25 s, and 2.94 T 1.14 s, respectively). No statistically significant differences were found between the DE group, Best CL, and the Worst CL subcategories in the average TFSQ Stability Time estimator. Fig. 3 shows the results of the average Best TFSQ for the four considered subcategories. The lowest values of the TFSQ estimator was achieved for the Bare Eye subcategory (105.68 T 6.03), and the average TFSQ was statistically significantly better (p = 0.03) than that of the DE group (114.80 T 12.44). Next, the average Best TFSQ was significantly better (p G 0.01) for the Best CL subcategory (118.10 T 7.94) than for that of the Worst CL (141.95 T 18.45). There was no statistically significant difference in the average Best TFSQ between the DE group and the Best CL subcategory. The group average RMS value was the highest for the Worst CL subcategory (9.65 T 0.90) and the lowest for the Bare Eye subcategory (7.37 T 1.80). Similarly, the estimated Pearson correlation coefficient was the highest (0.85 T 0.14) and the lowest (0.52 T 0.27) for the Worst CL and Bare Eye subcategories, respectively. The latter indicates that the proposed model is better suited for the CL tear film dynamics than for the bare eye scenarios. The former simply means that the variability of TFSQ in Bare Eye measurements is lower than those in CL acquisitions. The differences in the early changes in the poststability phase within the first 4 s are presented in Fig. 4. The AUC increases similarly for both the Best CL and the Worst CL subcategories. The slope of a linear approximation to AUC is 7.82 and 6.91 for Best CL and Worst CL subcategories, respectively. A substantially smaller change was achieved for the DE group with a linear slope of 4.96. This indicates that the tear film behaves better on the physiologic surface of the DE than on the artificial surface of a CL. Note that, for most of the normal eyes, the first power function was fitted to almost the complete sequence of data as indicated in Fig. 2A; therefore, this group was not taken into consideration in the analysis of AUC.
DISCUSSION
FIGURE 2. Examples of fitting the iterative power-exponential model to data recorded on bare eye with normal tear film (A), the best contact lens material (B), the worst contact lens material (C), and a dry eye (D).
Lateral shearing interferometry has the capability to observe subtle changes in the tear film surface as evidenced in our earlier studies. This technique of measurement performed well in distinguishing DEs from those having a normal tear film,13 and it showed the ability to differentiate CL materials through the analysis of the average TFSQ in natural blinking conditions.18 When the TFSQ is measured in suppressed blinking conditions on a CL, its dynamics essentially follow a two-phase process in contrast to bare eye or DE measurements that conform, in most cases, to the previously suggested multilinear model consisting of up to five distinct phases.4 A new model has been proposed to describe the two-phase process of tear film dynamics on contact lenses. The first phase of this process is well modeled by a power function, whereas the second phase is better fitted by an exponential function that is
Optometry and Vision Science, Vol. 91, No. 12, December 2014
Copyright © American Academy of Optometry. Unauthorized reproduction of this article is prohibited.
1410 Tear Film Dynamics on Soft Contact LensesVSzczesna-Iskander and Iskander TABLE 1.
Comparison of four parameters received from the proposed model of TFSQ behavior analyzed in four conditions: on the Best CL, Worst CL, DE, and normal eye Parameter
Eye
Best CL
Worst CL
DE
2.94 T 1.14 2.99
3.95 T 2.87 4.16
141.95 T 18.45 148.57
114.80 T 12.44 110.08
TFSQ stability time
Mean T SD Median
11.95 T 5.63 11.43
4.28 T 2.25 3.65
Best TFSQ
Mean T SD Median
105.68 T 6.03 104.90
118.10 T 7.94 120.73
Min RMS
Mean T SD Median
7.37 T 1.80 6.19
8.76 T 1.31 8.99
9.65 T 0.90 9.68
9.63 T 1.88 9.35
R2
Mean T SD Median
0.52 T 0.27 0.56
0.80 T 0.17 0.90
0.85 T 0.14 0.90
0.75 T 0.18 0.77
TFSQ, Tear Film Surface Quality; RMS, root mean square; R 2, coefficient of determination; Best CL, the best contact lens material; Worst CL, the worst contact lens material; DE, dry eye; Eye, bare eye.
conditioned on the first power function. The two types of functions were chosen empirically and found to be better suited to the tear film dynamics on CLs than to those of the bare eye and DE. In the majority of the recorded interferogram sequences of prelens TFSQ, a stage is observed, in which the estimated TFSQ abruptly changes. This suggests that, after a blink, the tear film on CL quickly smooths out to achieve the best possible (materialdependent) TFSQ and then follows a phase of sudden deterioration. We attribute the phenomenon of a sudden deterioration phase of the tear film on the CL to the wetting characteristics of the lens27 and particularly to the process of dewetting.15 Note that, in some of the measurements, the tear film deterioration phase caused the disappearance of interferogram fringe information, suggesting complete dryness of the lens surface. Such a phase is not normally observed in the case of bare eye measurements (both normal eyes and DEs) where the tear film thinning phase is much more subtle and more defined. It is suggested that the precorneal behavior of the tear film thinning is mainly caused by the evaporation but also related to the tangential flow surface tension gradients.3,15,17 Based on the previously described experiment,18 where four daily disposable CLs were tested, the best and worst lens
measurements acquired in natural blinking conditions (that caused the least and highest declines in TFSQ, respectively) were chosen for each of the subjects. The differences in tear film quality between those lens materials were confirmed by the results of group statistics of parameters extracted from the proposed model. The average Best TFSQ was significantly better for the Best CL than for the Worst CL. The time of TFSQ stability was also longer for the Best CL; however, the difference was not statistically significant. Shorter stability time of the tear film on contact lenses, rather than on the bare eye, confirmed earlier observations.28,29 The particular type of the lens material did play some role in the first phase of TFSQ dynamics affecting both the TFSQ Stability Time and the Best TFSQ. However, regardless of the initial interaction between lens material and tear film, the study showed that, when destabilization of the tear film surface starts, it progresses similarly quickly on the Best and the Worst CLs. This suggests that the material properties of a CL have little influence on tear film dynamics after the process of dewetting occurs. It is worth noting that the transition point from the first to the second phase of TFSQ dynamics was on average less than 3 s for the Worst CL and just above 4 s for the Best CLVvalues similar to those achieved for the DE group and shorter than an average interblink interval.30 Developing contact lens materials that would delay this
FIGURE 3.
FIGURE 4.
The best TFSQ_Av in four groups of ocular conditions. Statistically significant differences are in bold font and marked with asterisks. Error bars indicate 1 SD.
The dynamics of TFSQ in the first 4 s after the intersection point illustrates the progressive tear film thinning process.
Optometry and Vision Science, Vol. 91, No. 12, December 2014
Copyright © American Academy of Optometry. Unauthorized reproduction of this article is prohibited.
Tear Film Dynamics on Soft Contact LensesVSzczesna-Iskander and Iskander
transition point could result in the future in a better quality of the prelens tear film.
ACKNOWLEDGMENTS Supported by the European Regional Development Fund within Innovative Economy Operational Program cofinanced by the Foundation for Polish Science (POMOST/2012-5/8/0072 to DS-I). Neither author has anything to disclose. Received May 2, 2014; accepted July 29, 2014.
REFERENCES 1. Tomlinson A, Khanal S. Assessment of tear film dynamics: quantification approach. Ocul Surf 2005;3:81Y95. 2. Tsubota K. Tear dynamics and dry eye. Prog Retin Eye Res 1998; 17:565Y96. 3. King-Smith PE, Nichols JJ, Nichols KK, Fink BA, Braun RJ. Contributions of evaporation and other mechanisms to tear film thinning and break-up. Optom Vis Sci 2008;85:623Y30. 4. Szczesna DH, Iskander DR. Lateral shearing interferometry for analysis of tear film surface kinetics. Optom Vis Sci 2010;87:513Y7. 5. Li KY, Yoon G. Changes in aberrations and retinal image quality due to tear film dynamics. Opt Express 2006;14:12552Y9. 6. Koh S, Maeda N. Wavefront sensing and the dynamics of tear film. Cornea 2007;26:S41Y5. 7. Maissa C, Guillon M. Tear film dynamics and lipid layer characteristics– effect of age and gender. Cont Lens Anterior Eye 2010;33:176Y82. 8. Ozdemir M, Temizdemir H. Age- and gender-related tear function changes in normal population. Eye (Lond) 2010;24:79Y83. 9. Jones MB, Please CP, McElwain DL, Fulford GR, Roberts AP, Collins MJ. Dynamics of tear film deposition and draining. Math Med Biol 2005;22:265Y88. 10. Braun RJ. Dynamics of the tear film. Annu Rev Fluid Mech 2012; 44:267Y97. 11. Braun RJ, Gewecke NR, Begley CG, King-Smith PE, Siddique JI. A model for tear film thinning with osmolarity and fluorescein. Invest Ophthalmol Vis Sci 2014;55:1133Y42. 12. Monte´s-Mico´ R1, Alio´ JL, Charman WN. Dynamic changes in the tear film in dry eyes. Invest Ophthalmol Vis Sci 2005;46:1615Y9. 13. Szczesna DH, Alonso-Caneiro D, Iskander DR, Read SA, Collins MJ. Predicting dry eye using noninvasive techniques of tear film surface assessment. Invest Ophthalmol Vis Sci 2011;52:751Y6. 14. Bruce AS, Mainstone JC, Golding TR. Analysis of tear film breakup on Etafilcon A hydrogel lenses. Biomaterials 2001;22:3249Y56. 15. Nichols JJ, Mitchell GL, King-Smith PE. Thinning rate of the precorneal and prelens tear films. Invest Ophthalmol Vis Sci 2005;46:2353Y61.
1411
16. Liu H, Thibos L, Begley CG, Bradley A. Measurement of the time course of optical quality and visual deterioration during tear break-up. Invest Ophthalmol Vis Sci 2010;51:3318Y26. 17. King-Smith PE, Hinel EA, Nichols JJ. Application of a novel interferometric method to investigate the relation between lipid layer thickness and tear film thinning. Invest Ophthalmol Vis Sci 2010; 51:2418Y23. 18. Szczesna-Iskander DH, Iskander DR, Read SA, Alonso-Caneiro D. Noninvasive in vivo assessment of soft contact lens type on tear film surface quality. Invest Ophthalmol Vis Sci 2012;53:525Y31. 19. Guillon JP. Tear film structure and contact lens. In: Holly FJ, Lamberts DW, MacKeen DL, eds. The Preocular Tear Film in Health, Disease, and Contact Lens Wear. Lubbock, TX: Dry Eye Institute; 1986:914Y39. 20. Guillon M, Maissa C, Girard-Claudon K, Cooper P. Influence of the tear film composition on tear film structure and symptomatology of soft contact lens wearers. Adv Exp Med Biol 2002;506:895Y9. 21. Korb DR. Tear film-contact lens interactions. Adv Exp Med Biol 1994;350:403Y10. 22. Rohit A, Willcox M, Stapleton F. Tear lipid layer and contact lens comfort: a review. Eye Contact Lens 2013;39:247Y53. 23. Varikooty J, Keir N, Simpson T. Estimating tear film spread and stability through tear hydrodynamics. Optom Vis Sci 2012;89: E1119Y24. 24. Karapanagiotis I, Gerberich WW. Polymer film rupturing in comparison with leveling and dewetting. Surf Sci 2005;594:192Y202. 25. McMonnies CW. Key questions in a dry eye history. J Am Optom Assoc 1986;57:512Y7. 26. Szczesna DH, Iskander DR. Robust estimation of tear film surface quality in lateral shearing interferometry. J Biomed Opt 2009; 14:064039. 27. Fagehi R, Tomlinson A, Manahilov V, Haddad M. Contact lens in vitro wettability by interferometry measures of drying dynamics. Eye Contact Lens 2013;39:365Y75. 28. Fonn D, Situ P, Simpson T. Hydrogel lens dehydration and subjective comfort and dryness ratings in symptomatic and asymptomatic contact lens wearers. Optom Vis Sci 1999;76:700Y4. 29. Guillon M, Maissa C. Contact lens wear affects tear film evaporation. Eye Contact Lens 2008;34:326Y30. 30. Johnston PR, Rodriguez J, Lane KJ, Ousler G, Abelson MB. The interblink interval in normal and dry eye subjects. Clin Ophthalmol 2013;7:253Y9.
Dorota Szczesna-Iskander Wroclaw University of Technology Wybrzeze Wyspianskiego 27 50-370 Wroclaw Poland e-mail: [email protected]
Optometry and Vision Science, Vol. 91, No. 12, December 2014
Copyright © American Academy of Optometry. Unauthorized reproduction of this article is prohibited.
