Purchase this article with an account.
P. King-Smith, E. Hinel, J. J. Nichols; Measurement of the Thickness of the Lipid Layer of the Tear Film Using Reflection Spectra. Invest. Ophthalmol. Vis. Sci. 2008;49(13):1540. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Measurement of lipid thickness is important for understanding its role in reducing evaporation. The reflection spectrum from the lipid layer depends on optical interference between reflections from its front and back, and hence can be used to estimate lipid thickness.
In our previous lipid thickness estimates (Optom Vis Sci 76, 19), reflection spectra were measured on a "relative" scale; any misalignment of the eye reduced the light reaching the spectroscope, so that absolute reflectance was unknown. Lipid thickness was estimated from the slope and curvature of the reflection spectrum. That analysis had limitations, e.g., lipid thickness of 50 and 60 nm gave almost identical "relative" spectra even though absolute reflectances were considerably different. Thus the optical system was modified to measure absolute reflectance. A narrow illumination cone is focussed to a small point (33 um) on the tear film, and this point is refocussed totally within the entrance slit of the spectroscope, so that, if the eye is not badly misaligned, all the reflected light passes within the measuring optical system. Reflectance spectra covered the range 550 to 1060 nm. Two recordings of 200 spectra each, over a 20 second period, were made from each of 33 normal subjects (15 female, mean age 35).
Results were analyzed in two ways. Assuming that lipid thickness within the measurement spot was uniform, lipid thickness (n=66) ranged from 15 to 158 nm with a median of 36 nm and a mean of 43 ± 25 (SD) nm. Good correlation was obtained between the first and second estimates, (Spearman R=0.82, P<0.0005). Better fits could be obtained by assuming that lipid thickness varied within the measurement spot (gaussian distribution of log thickness); in that case, median thickness within the measurement spot ranged from 6 to 185 nm with a median of 28 nm and a mean of 37 ± 32 nm.
Our results agree well with those of Olsen (Acta Ophthalmol 63, 428) who measured absolute reflectance at two wavelengths, but are thinner than more recent values based on color appearance by E Goto, DR Korb and others. A wide range of lipid thickness was observed and thinner layers may be only a few molecules thick. We speculate that when the lipid layer is so thin, there may be insufficient polar lipids to form a good interface with the aqueous tears, thus disturbing the structure of the lipid layer and its barrier resistance to evaporation.
This PDF is available to Subscribers Only