If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Herbert Wertheim School of Optometry & Vision Science, University of California, Berkeley, CA 94720, United StatesClinical Research Center, Herbert Wertheim School of Optometry & Vision Science, University of California, Berkeley, CA 94720, United StatesChemical and Biomolecular Engineering Department, University of California, Berkeley, CA 94720, United States
Herbert Wertheim School of Optometry & Vision Science, University of California, Berkeley, CA 94720, United StatesClinical Research Center, Herbert Wertheim School of Optometry & Vision Science, University of California, Berkeley, CA 94720, United States
Clinical Research Center, Herbert Wertheim School of Optometry & Vision Science, University of California, Berkeley, CA 94720, United StatesChemical and Biomolecular Engineering Department, University of California, Berkeley, CA 94720, United States
Clinical Research Center, Herbert Wertheim School of Optometry & Vision Science, University of California, Berkeley, CA 94720, United StatesChemical and Biomolecular Engineering Department, University of California, Berkeley, CA 94720, United StatesCooperVision Inc., Pleasanton, CA 94588, United States
Herbert Wertheim School of Optometry & Vision Science, University of California, Berkeley, CA 94720, United StatesChemical and Biomolecular Engineering Department, University of California, Berkeley, CA 94720, United States
Herbert Wertheim School of Optometry & Vision Science, University of California, Berkeley, CA 94720, United StatesClinical Research Center, Herbert Wertheim School of Optometry & Vision Science, University of California, Berkeley, CA 94720, United States
With soft-contact-lens wear, evaporation of the pre-lens tear film affects the osmolarity of the post-lens tear film and this can introduce a hyperosmotic environment at the corneal epithelium, leading to discomfort. The purposes of the study are to ascertain whether there are differences in evaporation flux (i.e., the evaporation rate per unit area) between symptomatic and asymptomatic soft-contact-lens wearers, to assess the repeatability of a flow evaporimeter, and to assess the relationship between evaporation fluxes, tear properties, and environmental conditions.
Methods
Closed-chamber evaporimeters commonly used in ocular-surface research do not control relative humidity and airflow, and, therefore, misestimate the actual tear-evaporation flux. A recently developed flow evaporimeter overcomes these limitations and was used to measure accurate in-vivo tear-evaporation fluxes with and without soft-contact-lens wear for symptomatic and asymptomatic habitual contact-lens wearers. Concomitantly, lipid-layer thickness, ocular-surface-temperature decline rate (i.e., °C/s), non-invasive tear break-up time, tear-meniscus height, Schirmer tear test, and environmental conditions were measured in a 5 visit study.
Results
Twenty-one symptomatic and 21 asymptomatic soft-contact-lens wearers completed the study. A thicker lipid layer was associated with slower evaporation flux (p < 0.001); higher evaporation flux was associated with faster tear breakup irrespective of lens wear (p = 0.006). Higher evaporation flux was also associated with faster ocular-surface-temperature decline rate (p < 0.001). Symptomatic lens wearers exhibited higher evaporation flux than did asymptomatic lens wearers, however, the results did not reach statistical significance (p = 0.053). Evaporation flux with lens wear was higher than without lens wear but was also not statistically significant (p = 0.110).
Conclusions
The repeatability of the Berkeley flow evaporimeter, associations between tear characteristics and evaporation flux, sample-size estimates, and near statistical significance in tear-evaporation flux between symptomatic and asymptomatic lens wearers all suggest that with sufficient sample sizes, the flow evaporimeter is a viable research tool to understand soft-contact-lens wear comfort.
]. Increased tear-evaporation rate is widely accepted to be caused by a suboptimal tear lipid layer, a common source of which is meibomian gland dysfunction [
] found increased tear-film evaporation rates with SCL wear relative to no SCL wear. An increase in evaporation rate theoretically increases the hyperosmolarity of the tear and may trigger lens-wear ocular discomfort that shares the same etiology as dry-eye discomfort during no-lens wear [
]. However, unlike the case of no-lens wear, the association between evaporation rate and ocular comfort or measured osmolarity has not been established [
] is because tear-meniscus osmolarity, which is where the osmolarity is typically measured, does not correspond to the osmolarity of the post-lens tear film, which is the tear film in direct contact with the cornea. Kim et al. [
] went on to show that increased pre-lens tear-film evaporation rates result in increased post-lens tear-film osmolarity and that salt diffusivity of the SCL is important in regulating the osmolarity of the post-lens tear film.
Evaporimeters used in most tear-evaporation studies are closed chambers, meaning that there is minimal-to-no airflow during the measurement of the humidity within the apparatus [
]. These types of evaporimeters result in significant error in evaporation-rate measurements due to non-uniform distribution of water–vapor concentration within the chamber and error in relative humidity measurements [
]. Further, as the human eye experiences various airflows throughout a given day (e.g., walking or running on a windy or windless day) closed-chamber evaporimeters do not provide evaporation flux (i.e., evaporation rate per unit area) or a rate that is representative of what human eyes actually experience throughout the day [
]. A flow evaporimeter, which provides controlled airflow and inlet humidity within the apparatus while measuring the relative humidity and temperature of both apparatus-inlet and apparatus-outlet airflows [
], their instrument does not allow setting airflow quantitatively. Moreover, the study of Tomlinson and Cedarstaff involved only 5 subjects, precluding rigorous statistical analysis [
] also developed a flow evaporimeter, but their instrument provides evaporation rate only at zero relative humidity, which is not characteristic of the human eye. To overcome these limitations, Peng et al. [
], tear evaporation flux for symptomatic (SYM) and asymptomatic (ASYM) SCL wearers are quantified during lens wear and during no-lens wear. Further, the influence of lipid-layer thickness on the evaporation flux and the effect of evaporation flux on non-invasive tear break-up time (NITBUT), tear meniscus height (TMH), ocular-surface-temperature decline (OSTD) rate, and Schirmer Tear Test (STT) strip wetted length are assessed. Finally, differences between measured evaporation fluxes between SYM and ASYM contact-lens wearers, the repeatability of the flow evaporimeter, and sample size estimate for future studies are also evaluated.
2. Material and methods
2.1 Berkeley flow evaporimeter
Details on the flow evaporimeter and measurement methodology are provided by Peng et al. [
]. Only information specific to this study is provided here along with some relevant details of the flow evaporimeter. For all evaporimeter measurements, the apparatus inlet air velocity and relative humidity of the evaporimeter were set as 8.5 cm/s and 40%, respectively. This airflow mimics sitting/standing in a typical air-conditioned room with normal ventilation [
]. Subjects were asked to maintain a comfortable stare into the far end of the evaporimeter while blinking every 3 s, regulated with metronome signals, until the evaporimeter outlet airflow humidity and temperature reached steady state. Then subjects were instructed to keep their eyes closed until the exit air humidity and temperature reached a new steady state. Time to steady state for open and closed eyes varies with the subject but is typically reached within 5 – 7 min [
]. The difference of the two evaporation rates excludes evaporation occurring from the skin and provides a correct tear evaporation rate. Following the image-analysis method of Peng et al. [
] palpebral-aperture area was determined for each subject; tear evaporation flux was determined by dividing the tear evaporation rate by the palpebral area. Each evaporation measurement took approximately 10–15 min to complete.
2.2 Study subjects
Subjects for the study were recruited from the University of California, Berkeley Clinical Research Center database. All subjects filled out a questionnaire regarding their ocular health and contact-lens history to ensure that they met all inclusion and exclusion criteria. To be eligible for the study, subjects must have had an oculo-visual examination within two years at the start of the study, be at least 18 years old in age, be free of any ocular disease or allergies, have worn contact lenses in the past, and have at least 20/40 or better visual acuity with their habitual vision correction. Subjects were also required to discontinue contact-lens wear at least 24 h before each of their study visits and to not participate in any other concurrent clinical studies. Written informed consent, with a complete description of the study procedures, goals, risks, and benefits was obtained from all subjects. This study observed the tenets of the Declaration of Helsinki and was approved by the University of California, Berkeley Committee for Protection of Human Subjects. Data safety and subject anonymity were adhered to in accordance with the Health Insurance Portability and Accountability Act (HIPAA).
Each ASYM SCL wearer was matched with a SYM SCL wearer for this study and each pair was required to be within ± 5 years in age and to be of the same gender and ethnicity (Asian versus Non-Asian). In this study, Asians are defined as Chinese, Korean, Taiwanese, Vietnamese, Pacific Islanders, and Japanese since they are shown to have different response to contact-lens wear than Non-Asians [
]. In addition, each subject pair’s total time of SCL-wear history had to be within ± 5 years of each other. Subjects were considered to be ASYM contact-lens wearers if they scored 1 or 2 on the Berkeley Dry Eye Flow Chart (DEFC) [
]. If subjects scored 3 on the Berkeley DEFC, they were ineligible to participate in the study.
2.3 Study protocol
All subjects participated in five visits: the first visit was to determine whether the subject was a SYM or ASYM lens wearer and qualified for the study. The subsequent two visits were to conduct evaporation and tear-film measurements without SCL wear, and the final two visits were to perform evaporation and tear-film measurements with SCL wear. Subjects were asked to awaken at least 2 h before each study visit, maintain a similar visit time for every visit (i.e., ± 1 h), and to start each visit before 10 in the morning (AM) to obtain intraday AM and afternoon (PM) measurements. Each visit was scheduled at least one week apart from the previous visit to ensure that the procedures of the previous visit did not affect subsequent measurements.
At the beginning of each study-measurement visit, subject wake-up time, visit time, hours of sleep, hours of near work performed within 24 h prior to visit, outside-environment temperature, outside-environment humidity, exam-room temperature, and exam-room humidity were documented. Following completion of the questionnaire and acclimation to the exam-room environment for at least 15 min, ocular-surface health assessments were conducted with slit-lamp biomicroscopy (SL120, Carl Zeiss Meditec Inc., Jena, Germany) under white light to ensure no active or pre-existing ocular pathology (e.g., superficial punctate keratitis, corneal scars, or infiltrates).
After the ocular-surface health assessment, subject’s right-eye lipid-layer thickness was assessed using the TearScience® LipiView® (Johnson and Johnson Vision Care, Inc., Jacksonville, FL). Subsequently, right-eye TMH was measured with an OCULUS Keratograph 5 M (Oculus, Inc., Arlington, WA) and NITBUT was measured with a Medmont Corneal Topographer (Medmont C-100, Medmont International Pty Ltd, Vermont, Australia). NITBUT was measured three times and averaged. Using a FLIR a655sc infrared thermographer (Teledyne FLIR LLC, Wilsonville, OR), OSTD rate in °C/s for the right eye was then determined following the procedures described in Li et al [
For the no-lens visits 2 and 3, evaporation flux of the right eye was determined using the flow evaporimeter immediately after taking the tear-film and thermal measurements. Following the evaporimeter measurement, subjects were asked to return 6 h later to obtain the PM TMH, NITBUT, OSTD rate, evaporation flux, and STT wetted length in the listed order. For STT wetted-length measurements, subjects were asked to close their eyes after strip (Alcon Laboratories Inc., Fort Worth, TX) insertion; the wetted length was measured after 5 min of strip insertion. No anesthetics were used for the STT. All environmental factors were measured again for the PM visit. Upon completion of the PM visit, ocular-surface health was reassessed with a slit lamp and fluorescein instillation to ensure that no adverse change had occurred on the ocular surface during the study.
For visits 4 and 5, subjects inserted Comfilcon A (CooperVision Inc., Pleasanton, CA) contact lenses after taking the lipid-layer measurements. Visual acuity of at least 20/40 was achieved with lens wear for all subjects by providing subjects with contact lenses that matched each subject’s prescription. Thirty min after lens insertion, the remaining tear measurements and AM evaporimeter measurement were taken for the right eye. Subjects were asked to return 6 h later to obtain the PM measurements. The 4 and 5 PM visit measurements consisted of the same measurements as PM visits 2 and 3 except that all measurements were conducted with contact lenses in situ. Upon completion of the PM visit, SCLs were removed and ocular-surface health was assessed with a slit lamp and fluorescein instillation.
2.4 Statistical analysis
Repeatability of the quantitative evaporation fluxes for no-lens wear and lens wear was assessed by examining exploratory plots and descriptive statistics, Limits of Agreement (LoA), Difference-vs.-Mean (DVM) plots [
], variance component analysis, and the intraclass correlation coefficient (ICC). Using the variance-component results and a range of other data-based variance estimates, the sample sizes required to detect a range of hypothetical group differences in tear-evaporation flux, with the standard 95% confidence and 80% power, were estimated.
Thorough exploratory analysis, examining descriptive statistics and graphs, distributions of variables, stratification of outcomes on study design variables (e.g., SYM/ASYM, AM/PM, no-lens wear/lens wear), and basic univariate statistical tests were also conducted. Then, linear mixed-effects modeling was employed to estimate the evaporation flux in SYM and ASYM subjects while taking into account potential within-subject correlations (i.e., between repeat measurements within visits, and over repeated visits), demographics and contact-lens history variables, and laboratory measurements. All models were examined based on F-test p-values, clinical relevance of effect sizes, and residual and other diagnostic plots.
3. Results
There were a total of 21 SYM and 21 ASYM SCL wearers that were matched and completed the study. One subject dropped out due to scheduling conflicts, requiring recruitment of one additional subject to pair successfully all SYM to ASYM lens wearers. Subject age ranged from 18 to 34 years with mean age (SD) of 23.5 (3.8) for the SYM group and 22.7 (3.8) for the ASYM group. Study demographics consisted of 86% female and 14% male with 52% of the study population being Asian and 48% being non-Asian. Although females had significantly higher evaporation flux (p < 0.001) than males (lens wear: 2.47 10-6 g/cm2/s; no lens wear: 2.69 10-6 g/cm2/s), the dearth of male subjects and the imbalance in male to female ratio warrants additional investigation regarding the effect of gender. Mean (SD) hours of near work within a 24 h period prior to the visit for the SYM and ASYM groups were 6.9 (2.8) h and 6.1 (3.4) h, respectively, and were not significantly different (p = 0.391). Similarly, the mean (SD) difference between the SYM and ASYM groups for years of habitual lens wear [SYM: 8.2 (3.9) years; ASYM: 7.7 (3.9) years; p = 0.640], lens spherical power [SYM: −3.93 (2.29) D; ASYM: −3.85 (1.59) D; p = 0.892], and lens cylindrical power [SYM: −0.15 (0.43) D; ASYM: −0.26 (0.56) D; p = 0.492] were not significantly different.
3.1 Evaporation flux
Descriptive statistics for evaporation flux are provided in Table 1. All univariate and stratified design models showed no statistically significant effect of measurement time (i.e., AM and PM) on evaporation flux (p > 0.371). Evaporation flux was greater with lens wear than without lens wear, although not significantly so (p = 0.110). SYM lens wearers also exhibited greater evaporation flux than ASYM lens wearers, however not significantly so (p = 0.053).
Table 1Descriptive Statistics for Quantified Evaporation Flux.
Group
Visit
Min
Max
Median
Mean
SD
95% CI
Evaporation Flux (10-6 g/cm2/s): No Contact Lens (Visits 2 and 3)
SYM
AM
0.76
5.70
3.72
3.61
1.32
3.21
4.01
ASYM
AM
0.91
6.51
2.84
3.25
1.45
2.81
3.69
SYM
PM
1.44
7.85
2.61
3.41
1.65
2.91
3.91
ASYM
PM
1.10
7.18
2.46
3.03
1.62
2.54
3.52
Evaporation Flux (10-6 g/cm2/s): Contact Lens (Visits 4 and 5)
External factors such as body temperature, exam-room temperature, exam-room humidity, outside-environment wind speed, outside-environment temperature, and outside-environment humidity were not significantly associated with the measured evaporimeter fluxes when the subjects were not wearing contact lenses.
Interestingly for SCL wear, outside-environment humidity was significantly related to evaporation flux (p = 0.013) with the observed range of outside-environment humidity (35–96%) corresponding to evaporation flux of up to when looking at the PM visit data only. This inverse relationship was driven by the SYM subjects, with ASYM subjects showing insignificant effect of outside-environment humidity on lens-wear evaporation flux. Similarly, outside-environment wind speed was significantly and directly related to evaporation flux (p = 0.001) in SYM subjects, with the observed 0–34 km/h range of wind speed corresponding to evaporation flux of up to when looking at the PM data only. Again, outside-environment wind speed did not have an effect on evaporation flux for ASYM subjects. AM wind-speed and outside-environment humidity data did not show any significant effect on the evaporation flux. No other external factors were significantly related to evaporation flux measured during lens wear.
3.2 Relationship between Tear-Film characteristics and evaporation flux
Descriptive statistics for TMH, NITBUT, OSTD rate, and STT strip wetted length are provided in Table 2. Descriptive statistics for average lipid-layer thickness, which were measured only in the morning and without lens wear, are provided in Table 3. Similar to evaporation flux, NITBUT (p = 0.127 lens wear, p = 0.893 no-lens wear) and meniscus height (p = 0.172 lens wear, p = 0.449 no-lens wear) were not significantly different between AM and PM for the no-lens-wear and lens-wear conditions. OSTD rates were also not significantly different between AM and PM for no-lens wear (p = 0.803) or for lens wear (p = 0.057). Since lipid-layer thickness and STT strip wetted lengths were only taken once per visit day, the effect of measurement time was not assessed.
Table 2Descriptive Statistics for Non-Invasive Tear Break-Up Time, Tear Meniscus Height, Ocular Surface Temperature Decline Rate, and Schirmer Test Wetted Length.
Group
Visit
Min
Max
Median
Mean
SD
95% CI
Non-Invasive Tear Break-Up Time (s): Visits 2 and 3 (No Contact Lens)
SYM
AM
4.23
21.50
8.00
8.65
3.92
7.47
9.84
ASYM
AM
2.87
86.93
9.05
12.86
14.96
8.33
17.38
SYM
PM
4.43
44.07
8.57
10.90
7.13
8.75
13.06
ASYM
PM
3.77
41.17
7.87
10.99
7.95
8.59
13.40
Non-Invasive Tear Break-Up Time (s): Visits 4 and 5 (Contact Lens)
SYM
AM
1.30
13.30
3.65
4.28
2.62
3.49
5.08
ASYM
AM
1.30
19.40
4.90
5.54
3.72
4.42
6.67
SYM
PM
0.80
18.90
4.50
5.29
3.42
4.26
6.33
ASYM
PM
1.80
32.60
4.90
6.51
6.26
4.62
8.41
Tear Meniscus Height (mm): Visits 2 and 3 (No Contact Lens)
SYM
AM
0.09
0.49
0.24
0.26
0.09
0.23
0.28
ASYM
AM
0.12
0.45
0.24
0.26
0.08
0.23
0.28
SYM
PM
0.12
0.49
0.24
0.26
0.08
0.23
0.28
ASYM
PM
0.12
0.52
0.23
0.24
0.07
0.22
0.26
Tear Meniscus Height (mm): Visits 4 and 5 (Contact Lens)
SYM
AM
0.11
0.51
0.23
0.24
0.07
0.22
0.26
ASYM
AM
0.10
0.41
0.24
0.23
0.07
0.21
0.25
SYM
PM
0.09
0.42
0.23
0.23
0.07
0.21
0.25
ASYM
PM
0.10
0.38
0.19
0.21
0.06
0.19
0.23
Ocular Surface Temperature Decline Rate (°C/s): Visits 2 and 3 (No Contact Lens)
SYM
AM
0.0057
0.2249
0.0805
0.0852
0.0497
0.0702
0.1002
ASYM
AM
0.0022
0.2703
0.0848
0.0837
0.0662
0.0637
0.1037
SYM
PM
0.0026
0.1956
0.0796
0.0809
0.0518
0.0653
0.0966
ASYM
PM
0.0073
0.2300
0.0798
0.0838
0.0556
0.0669
0.1006
Ocular Surface Temperature Decline Rate (°C/s): Visits 4 and 5 (Contact Lens)
SYM
AM
0.0119
0.2417
0.0916
0.0970
0.0578
0.0795
0.1145
ASYM
AM
0.0034
0.2216
0.0829
0.0923
0.0577
0.0749
0.1098
SYM
PM
0.0157
0.1899
0.0769
0.0813
0.0474
0.0670
0.0956
ASYM
PM
0.0053
0.2062
0.0680
0.0781
0.0564
0.0611
0.0952
Schirmer Test Wetted Length (mm): Visits 2 and 3 (No Contact Lens)
SYM
PM
0
30
17.0
16.86
10.44
13.70
20.01
ASYM
PM
2
30
20.5
19.79
8.84
17.11
22.46
Schirmer Test Wetted Length (mm): Visits 4 and 5 (Contact Lens)
Fig. 1 summarizes the relationships between measured tear characteristics and evaporation flux. Arrows indicate statistically significant associations from the best fitting multivariable model. A thicker lipid layer is significantly associated with lower evaporation flux (p < 0.001). In turn, lower evaporation flux is associated with longer NITBUT (p = 0.006), decreased OSTD rate (p < 0.001), and, surprisingly, shorter STT strip wetted length (p = 0.033) (see Discussion). As expected, TMH was not associated with evaporation flux. SCL wear was significantly associated with shorter NITBUT, by an estimated 5.09 s (p = 0.028) compared with no lens wear. SYM lens wearers had 3.27 mm less SST strip wetted length than ASYM lens wearers (p = 0.025). Other tear-film characteristics were not significantly related to lens-wear symptomatology.
Fig. 1Mixed effects models of tear properties and evaporation flux. Arrows indicate statistically significant model effects. One unit change of the measurement made at the left of the arrow results in change in the measurement made at the right of the arrow by the value specified in black (e.g., 1 10−6 g/cm2/s change in the evaporation flux results in -0.78 s change in the NITBUT).
3.3 Repeatability and sample size analysis for evaporation flux
Mean inter-visit differences, LoA, and the ICC for no-lens wear and lens-wear evaporation fluxes are provided in Table 4. The mean inter-visit differences for the AM visits were 0.39 for no-contact-lens wear and −0.32 for contact-lens wear. Inter-visit agreement improved in the PM visits, with mean inter-visit differences of 0.19 for contact-lens wear and −0.14 for no-lens wear. The ICC for AM and PM for lens wear and no-lens wear all indicate that repeated evaporimeter measurements had moderate-to-good overall agreement [
]. With no-lens wear the ICC was 0.53 for both AM and PM visits. With contact-lens wear, the ICC showed somewhat better inter-visit agreement (0.74 in the AM, 0.60 in the PM). No noticeable trends in the dependence of the inter-visit differences on the magnitude of the measurement were observed in the DVM plots (not shown).
Table 4Repeatability of Evaporation Flux.
Visit
Mean-Difference (10-6 g/cm2/s)
Intraclass Correlation Coefficients
Limits of Agreement
p-value
Lower
Upper
No Contact Lens Evaporation Flux (Visit 2 – Visit 3)
Table 5 provides sample-size estimates for future group studies using the evaporimeter based on the true between-subject variance obtained from variance-component analysis. In addition to statistical confidence and power (set for this analysis at 95% and 80%, respectively), sample sizes depend upon the magnitude of the difference between groups one wishes to detect, and the variance of the measurement. Table 5 also presents sample-size estimates for a range of hypothetical group differences in evaporation flux, over a range of variance assumptions obtained from gathered experimental data. The sample size in the current study is large enough to detect a 20% difference in evaporation flux between groups if the true variance is close to the lower variance assumptions. With a more moderate variance assumption, a 10% difference in evaporation flux would be detectable with just under 200 subjects (∼100 subjects for each cohort). Larger sample sizes and better variance estimates in the future will improve sample-size projections. Note that Table 5 shows total sample sizes for a 2-group study; the most efficient study design would be to enroll half that number in each of the 2 study groups.
The repeatability of the Berkeley flow evaporimeter and the difference in tear-evaporation flux between symptomatic and asymptomatic lens wearers were assessed. Along with the evaporation flux, relationships between evaporation flux with various tear-film properties (i.e., lipid-layer thickness, tear-meniscus height, non-invasive tear break-up time, STT wetted length, and OSTD rate) were examined.
The flow evaporimeter exhibited moderate-to-good repeatability, with repeat measurements on individual eyes in the AM expected to fall within 0.39 10-6 g/cm2/s for no CL wear and within 0.32 10−6 g/cm2/s for CL wear, with 95% probability. Repeatability was better in the PM with repeat measurements expected to fall within 0.19 10−6 g/cm2/s for no CL wear and within 0.14 10−6 g/cm2/s for CL wear, with 95% probability. It is difficult to ascertain the clinical implications of these differences because the relationships between evaporation flux and important clinical outcomes have not been widely studied to date. We note that the inter-visit differences observed were an order of magnitude smaller than the mean evaporation flux values under all conditions in this study. While there is no absolute scale for interpreting the magnitude of the ICC and no standard qualitative terms for different values of the statistic, we interpret the repeatability of the flow evaporimeter to be moderate-to-good based on the two widely cited guidelines of Cicchetti [
]. In sum, the repeatability results suggest that although the flow evaporimeter is not currently sufficiently precise for individual evaporation flux monitoring in the clinical patient setting, it holds significant promise in the research setting, with relatively small differences between study groups detectable with manageable sample sizes (e.g., <200 eyes).
Lipid-layer thickness of the tear film is widely believed to regulate tear evaporation rate [
]. The other tear characteristics listed in Fig. 1 were hypothesized to depend on the tear evaporation flux. The statistically significant relationships between lipid-layer thickness and tear-evaporation flux, and between tear-evaporation flux and NITBUT reinforce the importance of the tear lipid layer in regulating tear thickness decline and, therefore, pre-corneal and pre-lens tear osmolarity during no-lens wear and lens wear, respectively. Since pre-corneal tear-film hyperosmolarity has been shown to be related to dry eye [
], results from this study provide additional evidence for the importance of the lipid-layer thickness on reducing dryness symptoms for both no-lens and lens wear. However, differences in evaporation flux between SYM and ASYM lens wearers did not reach statistical significance (p = 0.053) despite higher evaporation flux with SYM lens wearers. The small sample size of this exploratory study is a potential explanation. Also, it is important to point out that SYM lens wearers were still habitual lens wearers and, therefore, did not include intolerant lens wearers. If intolerant lens wearers were included in the study, differences between SYM and ASYM evaporation flux would likely have been greater. Moreover, the SYM and ASYM groups were determined from lens discomfort experienced with their habitual lenses. Due to differences in salt diffusivity among different contact lenses, there is a possibility that the ASYM lens-wearing group may have included some subjects that were protected from post-lens tear film hyperosmolarity by low-salt-diffusivity habitual contact lenses despite having high evaporation flux [
Despite results here showing non-statistically significant higher evaporation flux with lens wear than with no-lens wear, NITBUT was significantly faster with lens wear than without lens wear. This suggests that NITBUT is affected by factors other than evaporation flux. Other potential factors affecting NITBUT could be the lack of glycocalyx on the SCL surface [
]. As results here suggest that factors affecting NITBUT during lens wear and without lens wear are not the same, lens-wear NITBUT should not be directly compared with no-lens-wear NITBUT. This is supported by Graham and Lin [
], who performed a limited flow-evaporimeter study for 5 subjects, found a significant increase in evaporation rate with lens wear than with no-lens wear. In this study, evaporation flux was found to be higher with lens wear than without, although not significantly so (p = 0.110). Further investigation is necessary to understand whether evaporation flux between lens wear and no lens wear is different, and what the mechanisms of any difference may be.
As expected, OSTD rate increased with increased tear-evaporation flux. Although OSTD is due to multiple forms of ocular surface cooling (i.e., radiative heat loss, convective heat loss, and evaporative cooling [
]), results from this study indicate that thermographic assessment of the OSTD rate may be a viable tool to measure tear evaporation behavior non-invasively. No correlation between TMH and evaporation flux was expected and is explainable by the small surface area of the tear meniscus exposed to the environment. A small surface area means that the large volume of tears in the meniscus remains relatively unchanged during the interblink period and is minimally affected by evaporation. Interestingly, outside-environment humidity and outside-environment wind speed measured in the PM did have a significant effect on the PM tear evaporation flux in the lens-wear group but not in the no-lens-wear group. Even within the lens-wear group, the outside environment affected only the SYM lens wearers and not the ASYM lens wearers. The effect of outside-environment humidity and wind speed on tear-evaporation flux in the SYM lens-wear group suggests that the SCL and/or the tear film is affected by the environment with prolonged lens wear for SYM lens wearers. This environmental effect was apparent despite the subject eyes being introduced to a controlled environment condition of the evaporimeter chamber for 10–15 min during the evaporation flux measurements. Perhaps tear production (e.g., aqueous tear and lipid layer) or the SCL is irreversibility affected by the environment for SYM lens wearers during prolonged lens wear until lens removal. Further investigation is required to understand better the effect of environmental factors on the SCL and/or on the tear film during lens wear.
Results from this study show that STT strip wetted length and evaporation flux are significantly related. However, lack of anesthetic can introduce reflex tearing and not sheathing the Schirmer strip allows tear evaporation from the strip during the test, both of which can affect wetted lengths [
]. Although the association between STT strip wetted length and evaporation flux may be real, it is just as likely that this association comes from the uncontrolled reflex tearing and tear evaporation from the Schirmer strip. The relationship between evaporative flux and lacrimal tear production rate should be further investigated by utilizing the modified Schirmer tear test [
]. Thus, results in Table 3 do not include accurate lipid-layer thickness for subjects that have lipid-layer thicknesses greater than 100 nm. However, the inverse relation between lipid-layer thickness and evaporation flux seen in this study should not be affected by this limitation because subjects with lipid-layer thicknesses greater than 100 nm truncated at that value still showed lower evaporation flux on average. Improved measurement technology will provide a more complete range of human lipid-layer thickness.
Although the airflow was intended to be that of a typical air-conditioned room while standing/sitting still, some of the subjects complained of discomfort with the airflow speed during the evaporimeter measurements. This suggests that the controlled airflow velocity chosen for this study may have been higher than the characteristics of the intended environment. This, along with the limitation of closed-chamber evaporimeters resulting in underestimated tear evaporation fluxes [
], could explain why measured tear evaporation flux of this study is in the higher echelon of the tear evaporation-flux range determined from the meta-analysis of Tomlinson et al. [
] has moderate-to-good repeatability and suggest that lipid-layer thickness regulates the evaporation flux during lens wear and no-lens wear. Evaporation flux was shown to relate directly with OSTD rate and inversely with NITBUT, but not with TMH. Although evaporation flux was also shown to correlate with the STT strip wetted length, further investigation is needed with the modified STT [
] to determine the accuracy of the trend and the contribution of the STT’s limitations. The study results showed that SYM lens wearers have higher evaporation flux than that of ASYM lens wearers, but this result did not reach statistical significance (p = 0.053). However, statistical significance could potentially have been reached if the study included intolerant lens wearers in the SYM group, if lens wear symptoms of subjects were determined based on the discomfort they experienced while wearing the same SCLs rather than their habitual lenses, by creating multiple groups of SYM lens wearers based on the type of discomfort they experience, and/or by having a larger study cohort.
Statistically insignificant tear-evaporation-flux differences between lens wear and no-lens wear in this study was inconsistent with the results of existing limited studies [
] showed that change in the pre-lens tear-film evaporation rate affects the hyperosmolarity of the post-lens tear film and can lead to ocular discomfort [
]. Therefore, further investigation is required to ascertain whether evaporation flux is different with lens wear than without lens wear. Rigorously conducted previous in-vitro experiments [
] and moderate-to-good repeatability of the Berkeley flow evaporimeter suggest that accurate and repeatable evaporation-flux measurements can be determined with large enough sample size (∼200 eyes). Sample-size estimation, significant associations between various tear properties and evaporation flux, and near statistical significance in tear-evaporation flux between ASYM and SYM lens wearers from this study suggest that a well-designed flow evaporimeter is a viable tool to understand ocular discomfort with and without SCL wear.
6. Grant/Financial support
Roberta J. Smith Research Fund (MCL), UCB CRC Unrestricted Fund (MCL), CooperVision, Inc. (MCL).
Declaration of Competing Interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Cheng-Chun Peng is an employee of CooperVision Inc. Remaining authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
Financial support from Roberta J. Smith Foundation, UCB CRC Unrestricted Fund, and CooperVision, Inc.
References
Rolando M.
Refojo M.F.
Tear evaporimeter for measuring water evaporation rate from the tear film under controlled conditions in humans.
00054-1&id=gr1.jpg){kind=link}