STAHL, ULRIKE Dipl-AO; HO, ARTHUR PhD, FAAO; BRENT, GEOFFREY PhD; NADUVILATH, THOMAS PhD; STAPLETON, FIONA PhD, FAAO
Dry eye sensation and ocular discomfort are often reported by contact lens wearers,1–4 and cause many wearers to reduce their lens-wearing time or to completely discontinue contact lens wear.5–8 Several studies have been conducted to understand the underlying causes of contact lens induced dryness and have shown that contact lenses cause a change in tear film structure and stability, compromising the ability of tears to fulfill their various functions.9–12 It has been reported that during contact lens wear, lipid layer thickness and tear film break-up time significantly decrease,13,14 conceivably leading to an increase in tear film evaporation,15,16 lens deposition, and tear film osmolality.17–19
Increased tear film osmolality has been associated with dry eye disease20–22 and is considered to be a key diagnostic factor in this condition.16 Tear osmolality is a function of the rate of tear secretion, drainage, absorption, and evaporation and can be regarded as a single parameter of tear film dynamics.23 Hyperosmolality has been suggested as a cause of ocular discomfort,16,24 corneal and conjunctival changes in dry eye disease,21,25–27 and the release of proinflammatory mediators.28,29
Increased tear film osmolality has also been found in soft and hard contact lens daily wear and extended wear9,18,30,31 and has been associated with dry eye status during contact lens wear.31 Tear film hyperosmolality during contact lens wear may be a function of evaporation, reduced lacrimal secretion, and contact lens osmolality.27,32–34 Karkkainen showed an increase in contact lens osmolality during contact lens wear and postulated that elevated contact lens osmolality could contribute to tear film osmolality by producing an osmotic gradient.34
Recent studies evaluating soft contact lens osmolality in vitro using a vapor pressure osmometer have varied in their methodology and produced conflicting results. Pensyl and Benjamin35 found the osmolalities of soft contact lenses to be similar to their equilibrating solutions, with the accuracy of the results possibly depending on the thickness of the lenses. In contrast, Karkkainen36 found soft contact lenses to be significantly hypertonic compared with their packaging solution.
To date, no studies have determined the accuracy of contact lens osmolality measurements, or have investigated the factors that influence osmolality measurements with the vapor pressure osmometer. A better understanding of these factors may assist in verifying or developing methods to determine contact lens osmolality. This in turn may help to design studies that evaluate the effect of contact lens and tear film osmolality on ocular comfort and the ocular surface during contact lens wear and their interaction. The aim of this study was to investigate the influence of sample volume and of contact lens material on osmolality readings made with a vapor pressure osmometer.
Osmolality was measured using a Wescor 5520 Vapro Vapor Pressure Osmometer (Wescor, Logan, UT). The sensing element of the osmometer is a fine-wire thermocouple hygrometer that determines the dew point temperature depression of the loaded specimen. The dew point temperature depression is a function of solution vapor pressure that in turn is an indirect measurement of the osmolality. The measurement process takes 80 s, and the osmolality reading is displayed in mmol/kg.
Solutions used in this study were standards, provided by Wescor, with osmolalities of 100, 290, and 1000 mmol/kg, and phosphate buffered saline (PBS) (0.2 g KCl, 1.15 g Na2HPO4, 0.2 g KH2PO4, 8 g NaCl, 1 l distilled H2O). PBS solutions were initially produced as 10 times concentrations of isotonic saline (0.9% NaCl) and then diluted to the required osmolality, such as 320 and 500 mmol/kg. In the study investigating the influence of the contact lens material, PBS with 290 mmol/kg was used.
Four different contact lens types, one from each of the four FDA polymer groups, were chosen. Twelve contact lenses with a power of −1.00 D of each of the following materials were used: Lotrafilcon B (O2Optix, CIBA Vision Corp., Duluth, GA, group I), Nelfilcon A (Focus Dailies, CIBA Vision Corp., group II), Balafilcon A (Purevision, Bausch & Lomb, Rochester, NY, group III), and Etafilcon A (1 Day Acuvue, Johnson & Johnson Vision Care, Jacksonville, FL, group IV).
All fluid specimens were applied onto Whatman no. 1 filter paper (Whatman, England). The diameter chosen was dependent on the sample volume and the size of the sample holder. For volumes of 10 μl, large filter papers with a diameter of 7.0 mm, provided by Wescor, were used. Small filter papers with a diameter of 3.2 mm were used for volumes of 2.0 μl and smaller, produced according to the manufacturers' recommendations.37
Before measurements were taken, the osmometer was calibrated with the routine calibration procedure, using a sample volume of 10.0 μl as recommended by the manufacturer,37 and the cleanliness of the thermocouple was checked. The osmometer was cleaned as soon as the contamination level was 3, to ensure a good linearity and high accuracy, especially with small sample volumes. For measurements using smaller volumes than 10.0 μl, the osmometer was calibrated on the average of three consecutive assays using 290 mmol/kg. To test the influence of the sample volume and of the contact lens materials, the osmometer was calibrated with a sample volume of 0.8 μl. A volume of 0.8 μl was calculated as the average fluid amount of the lens discs of the chosen material. For all other measurements, the osmometer was calibrated with the intended sample size volume of the specimens.
To compensate for the effects of possible internal temperature changes, a full measurement cycle on a dry, empty chamber was carried out after every five measurements.
Accuracy of the Osmometer.
To verify the accuracy of osmolality measurements with the Wescor Vapor Pressure osmometer, a set of 20 measurements was carried out with each of the following sample volume and osmolality combinations. Sample volumes of 0.8, 2.0, and 10.0 μl were tested each with 290, 320, and 1000 mmol/kg.
Additionally, 20 measurements with a sample volume of 0.5 μl were carried out with 290 mmol/kg only, as other osmolalities are outside the measurement range of the osmometer for such small volumes.
Influence of Sample Volume.
The sample volume of lens discs can be expected to range from about 0.3 to 1.5 μl, with the mean volume of 0.8 μl, depending on the water content of the material. To test the influence of a difference between calibration and sample volume, sample volumes ranging from 0.5 to 1.1 μl in 0.1 steps were measured each 10 times. The order of the applied volumes was randomized. Measurements were carried out with solutions of 290, 320, 500, and 1000 mmol/kg for each sample volume.
Evaporation Rate in the Osmometer.
To measure the evaporation rate in the osmometer, the fluid amount without and after osmometry was compared. After weighing the small sample holder with dry filter paper on an electronic balance (AND ER-182, A&D Company, Tokyo, Japan), a sample volume of 2.0 μl was applied, and the sample holder with the wetted filter paper was weighed. The same procedure was repeated with carrying out an osmolality measurement cycle before weighing the wet filter paper. Both procedures were repeated each 10 times and carried out as quickly as possible.
Influence of Contact Lens Material.
To investigate the influence of different contact lens materials on osmolality readings, a 3.4-mm disc was trephined from the center of each contact lens. After equilibration in saline solution for 1 h, the lens discs were removed from their vials, blotted, and placed on a glass slide. After dehydration in an oven for 16 h at 105°C, the glass slides with the specimens were cooled in a desiccator to avoid any absorption of moisture from the surrounding air, and were then weighed on an electronic balance. The lens discs were then rehydrated in 1 ml of PBS 290 mmol/kg and allowed to equilibrate for 2 h in separate closed vials. PBS with 290 mmol/kg was chosen to allow comparison of the measurements against the respective standard after five measurements and to mimic the tear film osmolality (expected to range between 285 and 320 mmol/kg).
After removal from the vial and blotting, the lens discs were inserted into the small sample holder of the osmometer. The time between removal from the vial and insertion was maintained as constant and short as possible. It was ensured that the lens lay flat in the sample holder, and no liquid droplets or bubbles were visible on the surface. After the osmolality measurement was carried out, the lens disc was weighed with the sample holder on an electronic balance.
Accuracy of osmolality readings was analyzed by comparing the measured osmolality with the nominal values using one-sample t tests with Bonferroni correction. The association of the difference between nominal and measured osmolality with sample volume and nominal values was tested using Pearson correlation to establish the randomness of the bias.
The influence of the difference between calibration (0.8 μl) and sample volume on measured osmolality readings was analyzed by comparing the measured osmolalities for each sample volume against the measured osmolality with a sample volume of 0.8 μl, using one-way ANOVA with Bonferroni adjustment.
General linear models were used to analyze the association of sample volume with measured osmolality. Linear and nonlinear terms were tested for significance. R2 values were used to determine the best model.
The influence of contact lenses was analyzed by comparing the measured osmolality between lens groups, using Brown-Forsyth F test followed by post hoc comparison using Games-Howell correction. The association of measured osmolality with contact lens water content and sample volume was modeled using linear mixed models. These factors were entered into the model simultaneously. The linear mixed model accounted for the correlated data within each lens type. The model tested for the significance of the main effects and interactions. The main effects and interactions terms were in the final model only if they were significant. The estimated equation of the final model was then used to plot estimated values of osmolality for each lens type. Statistical significance was set at 5% for each test variable. SPSS for Windows version 14 (Chicago, IL) was used for data analysis.
Accuracy of the Osmometer.
Mean osmolalities are shown in Table 1. For each nominal, the readings were accurate (all p > 0.05, Table 1). Standard deviation increased with decrease in sample volume. The difference between nominal and measured osmolality did not correlate with sample volume (r = 0.002) and nominal osmolality (r = −0.133).
The osmolality measurement with 0.5 μl was not acceptable, as the mean difference to the nominal value of the solution was −3.4 mmol/kg (286.6 ± 5.7 mmol/kg, p = 0.015.)
Influence of Sample Volume.
Measured osmolalities varied with sample volume for all nominal solutions if the sample and calibration volume were not identical (Fig. 1, Table 2). Sample volumes larger than the calibration volume resulted in lower osmolality readings, whereas smaller sample volumes resulted in increased osmolality readings. Although the effect of volume on osmolality readings varied with nominal solution osmolalities; in general, sample volume differences >0.1 μl significantly affected osmolality readings (Table 2).
To eliminate the systematic error that is introduced by a difference in calibration and sample volume and thus correct measured osmolality readings, a regression equation was developed using general linear models. The equation to correct the measured osmolality readings is described as follows:
Corrected osmolality = −103.4 + measured osmolality + (194.4 × sample volume) + (−83.3 × sample volume2).
(r2 = 0.999, p < 0.001)
The measured osmolality of 331.8 related to a nominal osmolality of 290 and sample volume of 0.5 appears as an outlier for data analysis. When the analysis was performed without this observation, it was noted that the sample volume's quadratic term in the equation remained significant. The coefficient of the quadratic term changed from −83.3 to −51.5, and its corresponding p value changed from 0.004 to 0.005. There was no change in the overall R2. There was no valid reason to remove these data from the analysis.
It was observed that the average difference before applying this correction, which was 2.6 ± 13.6 mmol/kg, decreased to 0.0 ± 4.6 mmol/kg after the correction. This standard deviation was similar to that of the paired differences of the accuracy section.
The mean difference between the fluid amount on filter paper without and after osmometry was 0.09 μl (2.02 ± 0.04 and 1.93 ± 0.11 μl, respectively).
The measured osmolality varied with contact lens material (Fig. 2). Measured osmolalities of Lotrafilcon B (358.8 ± 45.4 mmol/kg) and Balafilcon A (356.7 ± 38.7 mmol/kg) were not significantly different to each other (post hoc p = 0.999). However, the measured osmolality of these lenses was significantly higher than that of Etafilcon A (298.2 ± 15.9 mmol/kg) and Nelfilcon A (281.2 ± 12.2 mmol/kg, post hoc p < 0.05 for each comparison). There was no statistical evidence to show a significant difference between the osmolalities of Etafilcon A and Nelfilcon A lenses (post hoc p = 0.056).
The measured osmolality varied also with sample volume. The association of these variables was best described using a combination of the linear term of sample volume (partial r = −0.57, p = 0.001) and its quadratic term (partial r = 0.41, p = 0.018), which accounted for 62.4% of the total variation (p < 0.001).
However, to model the association of water content and sample volume with measured osmolality, the mixed linear model was used to control the within-lens variance due to repeated lens samples. The factors associated with measured osmolality were best described by water content (p = 0.001), sample volume (p = 0.006), and their interaction (p = 0.024). Inclusion of the interaction term increased the R2 value from 66% to 71.6% (Fig. 3). The equation describing this association was:
Measured osmolality = 536 + (−3.4 × water content) + (−202.9 × sample volume) + (2.7 × water content × sample volume).
This study has demonstrated that osmolality readings with a vapor pressure osmometer vary with calibration and sample volume, and with different contact lens materials.
Sample volume is strongly associated with measured osmolality (r2 = 0.9999). Differences of more than 0.1 μl between the calibration and sample volume significantly affected osmolality readings, with sample volumes larger than the calibration volume resulting in lower readings than the nominal value of the solution, and smaller volumes in higher readings, independent of the nominal osmolality of the solution. We hypothesize that larger sample volumes are associated with a higher vapor pressure in the chamber, resulting in a lower osmolality reading due to a decreased dew point temperature depression.
When the sample and calibration volume are identical, a high accuracy and repeatability was obtained for sample volumes between 0.8 and 10.0 μl. These data and the low variance of measurements are in accordance with the data reported by Pensyl and Benjamin.38 In addition, the repeatability of the osmometer was dependent on the sample volume used, with the highest repeatability with the largest volume.38 In contrast to the data reported by Pensyl and Benjamin, measurement with a sample volume of 0.5 μl were significantly different to the nominal value of the solution and not acceptable. The standard deviation and the maximum difference between the measured and the nominal osmolality were similar,38 but a trend toward lower osmolality measurements occurred in the present study. Measurements with sample volumes as low as 0.5 μl are difficult to conduct, as the filter paper needs to be fully saturated and an increased sample size may help increasing the accuracy of the measurement.
Despite the dependence of osmolality readings on the sample volume, vapor pressure osmometry has advantages over freezing point depression osmometry. The measurement using vapor pressure osmometry avoids the alteration of the physical state of the specimen and therefore allows the measurement of complex specimens such as tissue samples.37 Therefore, it is technically possible to insert contact lenses into the osmometer and to measure their osmolality.
Direct measurement of equilibrated contact lens samples varied with the fluid volume associated with the lens disc and the lens material water content. This may explain the discrepancies between previous studies. In the present study, the measured osmolality of higher water content lenses (Etafilcon A, 58% and Nefilcon 69%) were similar to the osmolality of the equilibrating solution. This is consistent with an earlier publication in which the measured osmolality of lenses with water contents between 58% to 62% were comparable to the equilibrating solution.35 Our estimates of fluid volume for the specified lens diameter for these water contents would be 1.0 to 1.8 μl, which would closely approximate the assumed calibration volume of 2.0 μl.
One possible explanation for the impact of water content on measured osmolality may be that water content or the ratio of free-to-bound water in the material influences the time taken for equilibration in the chamber. Lower water content lenses or lenses with a lower free-to-bound water ratio may take longer to equilibrate and in standard measurement mode may result in a lower vapor pressure.
These two factors and their interaction account for 71.6% of the variance in measured osmolality. Other factors might include evaporation during the time taken for preparing the lens disc for measurement, lens thickness, other material characteristics such as lens modulus, surface treatment, and surface tension.
One major limitation of this direct lens measurement approach is inaccuracy in the measurement of lens fluid volume, in which multiple measurement steps are used. Small differences in final fluid volume estimate have a major impact on the final osmolality measurement. Other methodological issues include poor centration of lenses before trephination, resulting in lens thickness profile and volume variations. Trephination may lead to rough edges with an unpredictable impact on measured osmolality.
To mitigate these potential methodological confounders, the time taken for lens disc preparation was maintained at 7 s for all samples, lens thickness profile was maintained as parallel as possible by using a constant lens power of −1.00 DS, and samples were discarded if the trephination was not precise.
The Wescor Vapro vapor pressure osmometer is designed to measure sample volumes of 10.0 μl and can be altered to measure very low volumes (lower than 4.0 μl) when using a smaller sample holder and smaller filter papers. This study has shown that the osmometer gives accurate and repeatable measurements, using sample volumes as low as 0.8 μl. However, it is essential to calibrate the osmometer with the intended sample volume, as differences between the calibration and sample volume will impact the osmolality readings.
This study has described a direct method for measurement of contact lens osmolality. Sample volume and lens material water content have a major impact on measured osmolality; however, other methodological and material factors may also play a role. This method has relevance for the in vitro measurement of lens osmolality, particularly for high water content hydrogel lenses. Measurement of low water content lenses in vitro may be better achieved with measurements in “Process Delay Mode”; however, the applicability of this technique to worn lenses has yet to be determined. To use the direct measurement method to evaluate the osmolality of worn contact lenses accurately, we would recommend first measuring the osmolality of the contact lens material in vitro and adjusting the measurement procedure if necessary.
Vision CRC is an Australian Commonwealth funded cooperative research center and under its conditions of funding Vision CRC is required to commercialize its research. As part of that commercialization activity, Vision CRC receives royalty income from the sale of silicon hydrogel contact lenses sold by Bausch & Lomb and CIBA Vision. The Institute for Eye Research (IER) is a not for profit research corporation that is a core participant in Vision CRC and its employees are entitled to benefit from such royalties. U.S. was also supported by the University of New South Wales through an International Postgraduate Research Scholarship, the Contact Lens Society of Australia and the German Academic Exchange Service.
Vision CRC, Level 3
Rupert Myers Building North Wing
Gate 14, Barker Street
Kensington, NSW 2052, Australia
1. Begley CG, Caffery B, Nichols KK, Chalmers R. Responses of contact lens wearers to a dry eye survey. Optom Vis Sci 2000;77:40–6.
2. Guillon M, Maissa C. Dry eye symptomatology of soft contact lens wearers and nonwearers. Optom Vis Sci 2005;82:829–34.
3. Begley CG, Chalmers RL, Mitchell GL, Nichols KK, Caffery B, Simpson T, DuToit R, Portello J, Davis L. Characterization of ocular surface symptoms from optometric practices in North America. Cornea 2001;20:610–18.
4. Chalmers RL, Begley CG. Dryness symptoms among an unselected clinical population with and without contact lens wear. Cont Lens Anterior Eye 2006;29:25–30.
5. Weed K, Fonn D, Potvin R. Frequent replacement of soft contact lenses reduces complications: 2 year results. Optom Vis Sci 1993;70 (suppl):140.
6. Young G, Veys J, Pritchard N, Coleman S. A multi-centre study of lapsed contact lens wearers. Ophthalmic Physiol Opt 2002;22:516–27.
7. Pritchard N, Fonn D, Brazeau D. Discontinuation of contact lens wear: a survey. Int Contact Lens Clin 1999;26:157–62.
8. Nichols JJ, Mitchell GL, Nichols KK, Chalmers R, Begley C. The performance of the contact lens dry eye questionnaire as a screening survey for contact lens-related dry eye. Cornea 2002;21:469–75.
9. Farris RL. Tear analysis in contact lens wearers. CLAO J 1986;12:106–11.
10. Faber E, Golding TR, Lowe R, Brennan NA. Effect of hydrogel lens wear on tear film stability. Optom Vis Sci 1991;68:380–4.
11. Korb DR. Tear film-contact lens interactions. Adv Exp Med Biol 1994;350:403–10.
12. Glasson MJ, Stapleton F, Keay L, Willcox MD. The effect of short term contact lens wear on the tear film and ocular surface characteristics of tolerant and intolerant wearers. Cont Lens Anterior Eye 2006;29:41–7.
13. Thai LC, Tomlinson A, Doane MG. Effect of contact lens materials on tear physiology. Optom Vis Sci 2004;81:194–204.
14. du Toit R, Situ P, Simpson T, Fonn D. The effects of six months of contact lens wear on the tear film, ocular surfaces, and symptoms of presbyopes. Optom Vis Sci 2001;78:455–62.
15. Mathers W. Evaporation from the ocular surface. Exp Eye Res 2004;78:389–94.
16. Lemp MA. Report of the National Eye Institute/Industry workshop on clinical trials in dry eyes. CLAO J 1995;21:221–32.
17. Guillon JP, Guillon M. Tear film examination of the contact lens patient. Optician 1993;206:21–9.
18. Iskeleli G, Karakoc Y, Aydin O, Yetik H, Uslu H, Kizilkaya M. Comparison of tear-film osmolarity in different types of contact lenses. CLAO J 2002;28:174–6.
19. Cedarstaff TH, Tomlinson A. A comparative study of tear evaporation rates and water content of soft contact lenses. Am J Optom Physiol Opt 1983;60:167–74.
20. Gilbard JP, Farris RL, Santamaria J II. Osmolarity of tear microvolumes in keratoconjunctivitis sicca. Arch Ophthalmol 1978;96:677–81.
21. Gilbard JP, Farris RL. Tear osmolarity and ocular surface disease in keratoconjunctivitis sicca. Arch Ophthalmol 1979;97:1642–6.
22. Farris RL. Tear osmolarity—a new gold standard? Adv Exp Med Biol 1994;350:495–503.
23. Tomlinson A, Khanal S. Assessment of tear film dynamics: quantification approach. Ocul Surf 2005;3:81–95.
24. Gilbard JP. Luo and colleagues turn the lights back on … on dry eye. Eye Contact Lens 2005;31:135–6.
25. Gilbard JP, Carter JB, Sang DN, Refojo MF, Hanninen LA, Kenyon KR. Morphologic effect of hyperosmolarity on rabbit corneal epithelium. Ophthalmology 1984;91:1205–12.
26. Gilbard JP, Rossi SR, Heyda KG. Tear film and ocular surface changes after closure of the meibomian gland orifices in the rabbit. Ophthalmology 1989;96:1180–6.
27. Gilbard JP, Rossi SR, Gray KL, Hanninen LA, Kenyon KR. Tear film osmolarity and ocular surface disease in two rabbit models for keratoconjunctivitis sicca. Invest Ophthalmol Vis Sci 1988;29:374–8.
28. Li DQ, Luo L, Chen Z, Kim HS, Song XJ, Pflugfelder SC. JNK and ERK MAP kinases mediate induction of IL-1β, TNF-α and IL-8 following hyperosmolar stress in human limbal epithelial cells. Exp Eye Res 2006;82:588–96.
29. Luo L, Li DQ, Corrales RM, Pflugfelder SC. Hyperosmolar saline is a proinflammatory stress on the mouse ocular surface. Eye Contact Lens 2005;31:186–93.
30. Martin DK. Osmolality of the tear fluid in the contralateral eye during monocular contact lens wear. Acta Ophthalmol (Copenh) 1987;65:551–5.
31. Nichols JJ, Sinnott LT. Tear film, contact lens, and patient-related factors associated with contact lens-related dry eye. Invest Ophthalmol Vis Sci 2006;47:1319–28.
32. Gilbard JP, Dartt DA. Changes in rabbit lacrimal gland fluid osmolarity with flow rate. Invest Ophthalmol Vis Sci 1982;23:804–6.
33. Gilbard JP, Gray KL, Rossi SR. A proposed mechanism for increased tear-film osmolarity in contact lens wearers. Am J Ophthalmol 1986;102:505–7.
34. Karkkainen TR. Probing the hydrogel lens osmotic gradient. Optom Vis Sci 2003;80 (Suppl):6.
35. Pensyl CD, Benjamin WJ. Tonicity of salines absorbed by disposable soft contact lenses. Optom Vis Sci 1999;76 (Suppl):240.
36. Karkkainen TR. The osmolality of disposable soft contact lenses and storage solutions. Optom Vis Sci 2000;77 (Suppl):179.
37. Wescor Inc. 5520 VAPRO Vapor Pressure Osmometer User's Manual. Logan, UT: Wescor; 2002.
38. Pensyl CD, Benjamin WJ. Vapor pressure osmometry: minimum sample microvolumes. Acta Ophthalmol Scand 1999;77:27–30.