Mechanisms of Rubbing-Related Corneal Trauma in Keratoconus

Charles W. McMonnies, MSc

Purpose: Corneal scarring in keratoconus, which is observed prior to contact lens wear and in association with a chronic habit of abnormal rubbing, suggests keratocyte change to repair phenotype in response to rubbing trauma. Methods: This review examines known and putative mechanisms for rubbing-related corneal trauma and cone formation.

Responses to eye rubbing (and possible causal links) may include increased corneal temperature, epithelial thinning, increased concentrations of inflammatory mediators in the precorneal tears, abnormal enzyme activity, large intraocular pressure spikes, high hydrostatic tissue pressure, thixotropically reduced ground substance viscosity, temporary displacement of ground substance from the corneal apex, buckling and flexure of fibrils associated with waves of corneal indentation, biomechanically coupled curvature transfer to the cone apex, slippage between collagen fibrils at the cone apex, and changes to keratocytes due to mechanical trauma and/or high hydrostatic pressure, in addition to scar formation. Cone formation appears to depend on loss of shear strength and may be consequence of reduction in ground substance viscosity and glue function, which could allow the cornea to bend and yield to intraocular pressure.

For some forms of keratoconus, reduction in shear strength and cone-forming deformation may be responses to rubbing trauma. Some of the mechanisms for corneal rubbing trauma may be relevant to post-laser-assisted in situ keratomileusis ectasia or complications following other types of corneal surgery. There appear to be indications for the control of chronic habits of abnormal rubbing.

Key Words: keratoconus, keratectasia, cornea, rubbing (Cornea 2009;28:607–615)

Received for publication March 31, 2008; revision received December 4, 2008; accepted December 6, 2008.
From the School of Optometry and Vision Science, University of New South Wales, Kensington, Australia.
Presented at the Global Keratoconus Congress, Las Vegas, January 2008.
The author does not have any financial interests to declare in relation to this p
per. Correspondence: Adjunct Professor C. W. McMonnies, MSc, 77 Cliff Avenue, Northbridge, Australia 2063 (e-mail:
Copyright ? 2009 by Lippincott Williams & Wilkins


Chronic habits of abnormal rubbing (CHAR) are strongly associated with the development of keratoconus (KC). This association has been demonstrated by numerous case series.1–3 For example, a case control study of 120 subjects with KC involved assessment of potential risk factors, including atopy, family history, eye rubbing, and contact lens wear.4 In the univariate analysis, there were associations between KC and atopy, family history, and eye rubbing.4 However, in the multivariate analysis, only eye rubbing was still a significant predictor of KC.4 Two large studies of KC revealed that approximately 50% of subjects reported vigorous or abnormal rubbing of at least 1 eye.5,6 However, a strong association is not a sufficient condition for establishing a causal role.7 Observational case reports usually have a low rank for inferring support for a causal hypothesis.8 However, the control eye feature of histories of unilateral CHAR and associated corresponding unilateral KC appear to provide stronger evidence for a contributory rubbing causal hypothesis. 3,9 Consistent with a variable phenotype, multiple mechanisms for the development of KC have been identified or proposed. These include increased intraocular pressure (IOP)10–12 and increased hydrostatic pressure13; epithelialstromal interactions11,14 with involvement of inflammatory mediators and loss of keratocytes15; abnormal enzyme function and loss of collagen and/or ground substance16,17; anomalous keratocyte function and altered fibrillogenesis11,18 and/or altered proteoglycan production18–21; reduced tensile strength11,22; reduced shear strength19,22; collagen slippage in the cone19,20,22,23; and cone formation by biomechanically coupled curvature transfer.24–27 This reviewexamines evidence for causal links between rubbing-related corneal trauma and the development and/or progression of KC.


When the eye is closed, corneal temperature is raised because of the lack of evaporation from the ocular surface and because of the proximity of the circulatory warmth of the palpebral conjunctiva.28 The increase is higher with blepharospasm, 28 possibly because of the increased proximity of the vascular palpebral conjunctiva to the cornea due to the associated elimination of Kessing’s space under the tarsal region of the lid.29 Eye closure during rubbing may involve tight squeezing, although the proximity of the palpebral conjunctiva to the cornea is very high during rubbing, irrespective of the force of closure. A marked inflammatory palpebral conjunctival response has been observed in animal studies of rubbing.30,31 Experimental eye rubbing was found to cause a mild and transient increase in ocular itching, chemosis, and hyperemia in human subjects.32 Increases in corneal temperature may be greater during prolonged vigorous rubbing because of eye closure alone. Any increase might vary with the degree of hyperemia induced. However, any friction between the palpebral conjunctiva and cornea due to rubbing may help raise corneal temperature. After rubbing ceases, recovery of normal corneal temperature may be slowed while conjunctival hyperemia persists. Thermokeratoplasty threshold temperatures in excess of 50C are required to ‘‘shrink’’ corneal collagen,33 and rubbing-related temperature spikes are likely to be minor by comparison. However, any increase in corneal temperature that is associated with rubbing, and related conjunctival injection, may contribute to some of the other responses to rubbing that are discussed below. For example, collagenase activity could be upregulated during periods of rubbing-related temperature spikes. Also, rubbingrelated thixotropic reduction in ground substance viscosity may be accelerated by increased temperature.


Keratocyte apoptosis has been shown to be associated with corneal wounding.15 This finding gave rise to the hypothesis that the disappearance of keratocytes (due to an imbalance between keratocyte apoptosis and proliferation) may be the underlying pathophysiological mechanism for the development of KC.15 Corneal fibroblasts in KC have been shown to have four-fold greater numbers of interleukin-1 receptors compared with normal eyes.34 Gradual loss of keratocytes and any associated reduction in fibrillogenesis and/or the production of proteoglycans may contribute to loss of stromal mass.15 These changes might occur as a result of increased keratocyte sensitivity to interleukin-1 production.15 However, this mechanism for keratocyte loss in KC could be accelerated if corneal trauma resulted from contact lens wear or eye rubbing.15 This risk may be increased if a cornea is exposed to contact lens trauma, as well as the rubbing that occurs before contact lens insertion and/or after contact lens removal. Experimental slight rubbing for 10 seconds by non-contact-lens-wearing normal subjects, using one finger and a smooth circular movement, was repeated 30 times over a 30-minute period.35 Keratocyte density was found to be reduced, and the tears of these eyes were found to have high concentrations of inflammatory mediators (interleukin-8 and epithelial growth factor).35 It may be significant that vigorous rubbing was not required to induce these changes. The responses may be greater when more force is used and rubbing episodes are longer. Keratocyte density was also found to be reduced in unadapted normal subjects after 2 hours’ wear of both soft and rigid lenses.35 The mechanical stimulation of the corneal surface due to the presence of a contact lens or from rubbing, and the consequent release of inflammatory mediators, may account for the observed reduction in keratocyte density.35,36 Rubbing-related changes in keratocytes that reduce their visibility during confocal microscopy may have contributed to the finding of reduced keratocyte density. Morphological changes may alter cell reflectivity, for example. This possibility is examined below, when cell shape changes in response to increased hydrostatic pressure37 are discussed.


A role for corneal epithelium in KC development, which includes interaction with the stroma, has been suggested.14 A study of the effects of rubbing the eyes of rats indicated upper palpebral conjunctival mast cell degranulation, as well as increased neutrophil and macrophage numbers, over a 24-hour period.30 In a similar study in rabbits, a significant alteration to the upper palpebral conjunctiva was also found.31 However, the corneal epithelium remained intact and appeared to be more resistant to the trauma of eye rubbing.31,38 The epithelial thickness of normal human corneas was found to be reduced by 18.4%, both centrally and midperipherally, after 15 seconds of rubbing.39 The experimental rubbing was in a circular pattern with use of light to moderate force and the finger pad of an index finger.39 Recovery to baseline thickness took 15 to 30 minutes centrally and 30 to 45 minutes mid-peripherally.39 Rubbing-related epithelial thinning may include cell flattening, as well as displacement from the rubbed area of, for example, any of the following: cells, extracellular fluid, cytoplasm from any burst cells, and/or mucin.39 These findings suggest a potential for significant rubbing-related epithelial trauma. The anterior corneal mosaic,40 or Fischer-Schweitzer polygonal reflex,41 can be observed in a variety of circumstances, after instillation of fluorescein, using cobalt blue filtered light.40 In some eyes (e.g., keratoconus) the polygonal mosaic can be seen after blinking and with hard contact lens wear.40 However, in most normal eyes, eye rubbing40 or the gentle application of a warm compress41 is necessary to elicit the honeycomb mosaic pattern. The detail of the pattern is constant over time and appears to be a manifestation of a structural anatomical feature.40 The anterior-most lamellae attached to the posterior surface of Bowman’s layer give rise to the anterior mosaic.22 A series of convex ridges can be generated on the surface of Bowman’s layer when the cornea is indented.42 These ridges correspond to an arrangement of strap-like stromal bundles, which insert into Bowman’s layer.42 When the corneal fibrils are relaxed by indentation, these ridges define a ‘‘chicken wire’’ pattern, which is responsible for the anterior corneal mosaic.22,42 It is possible that the effects of rubbing forces are greater for the epithelium that is compressed over these ridges. As yet, however, there does not appear to be any evidence of epithelial trauma that could specifically be attributed to a Bowman’s membrane ridge– related model. For patients with KC, compressive forces on the corneal epithelium may be sustained for up to 5 minutes in a single rubbing episode.43 In addition, for KC subjects the use of one or more knuckles, a rotary, grinding motion, a force greater than 4.5 kg/2.54 cm2, and multiple episodes of prolonged rubbing each day are frequently involved.43 Although the squamous epithelial cells appear to be more resistant to this form of trauma in animal studies,30,31,38 wing and basal cells may be more susceptible to compressive mechanical trauma. This possibility is suggested by the finding of marked changes in the columnar epithelial cells of the palpebral conjunctiva in animal rubbing studies.30,31,38 However, studies of the effects of experimental rubbing in animals may not reflect the risk for the human cornea to develop KC. For example, when compared with apparently healthy rabbit or rat cornea, a human cornea that is predisposed to the development of KC may be more susceptible to rubbing-related epithelial injury. Studies of experimental rubbing in animals may need to be repeated more than once every day, over weeks or months, in order for investigators to determine the potential for adverse epithelial and other corneal tissue responses that are comparable with the CHAR of some KC subjects. The long duration of inflammatory responses to rubbing in animal studies30 indicates the possibility that lack of time to recover between CHAR episodes31 could result in cumulative tissue changes. Sustained higher levels of inflammatory mediators in the tears due to repeated rubbing episodes may be significant in KC development and progression. A case of inferior corneal thinning and high astigmatism, with some features of KC, was reported in association with chronic ocular rosacea.44 The inferior corneal thinning may have been related to chronic exposure of the inferior cornea to inflammatory and matrix-degrading factors in the inferior tear meniscus.44 Downgaze during reading or other form of near vision may expose the corneal apex to high concentrations of inflammatory and matrix-degrading factors that are retained in the inferior tear meniscus following a rubbing episode. The excess fluid in epithelial edema is primarily between cells, tending to separate them and increase fragility through loss of junctional contact, so that susceptibility to abrasion is increased.45 Rubbing that occurs immediately on waking or after contact lens removal (removal relief rubbing) may have the potential to cause greater corneal trauma if the epithelium is edematous (and/or otherwise compromised by lens wear).46 The observation that the cone commences most often in the paracentral inferonasal cornea, where the most senile epithelial cells may be more vulnerable to mechanical forces, might help link chronic rubbing-related trauma to cone formation.47 Rubbing rabbit eyes did not result in a significant change to the cell membrane phospholipids of the cornea,48 as might be expected if the integrity of the cell membranes had been disturbed by rubbing. However, there may be other markers for rubbing-induced epithelial mechanical trauma.49 Following a review of corneal wound healing, the conclusion was drawn that, regardless of the type of injury, there may be a common set of epithelial cell–keratocyte interactions, mediated by growth factors, chemokines, cytokines, and neural peptides.49 A fine balance between wound healing and undesired tissue destruction appears to be mediated by signaling components and factors such as matrix metalloproteinases and superoxide dismutase.49 For example, one signal for keratocytes to undergo fibrotic change may be the presence of transforming growth factor-(beta2) released from the corneal epithelium.49


The fine reticular scars of Bowman’s membrane tears are a well-known and characteristic feature of KC and are preceded by visible dehiscences at this level. These dehiscences are seen as fine, dark rents in an opalescent ground, when viewed with the broad oblique beam of a slit lamp.12 A simplistic interpretation is that the primary event is rupture of Bowman’s layer, whereas scarring is a secondary reaction of the injured tissue.12 More extensive and dense superficial scarring in KC may occur, even in eyes that have not worn a contact lens.12 When seen prior to contact lens wear and in combination with a history of a CHAR, this type of scarring may represent evidence of rubbing trauma. The percentage of 1209 patients enrolled in the Collaborative Longitudinal Evaluation of Keratoconus (CLEK) who did not wear contact lenses was 19.7%.5 Of these, 15% were scarred at baseline.50 Any analysis of an association between scarring in KC prior to contact lens wear and histories of CHAR has not been reported. However, wearing rigid lenses has been identified as a risk factor for developing scarring, and scarring increases with age.51 At the end of the 8-year CLEK study follow-up period, 20% of non-rigid lens wearers had developed scarring.51 Again, this finding suggests the possibility that rubbing-related trauma causes or contributes to scarring. However, corneal stromal scar tissue is markedly inelastic, and its mechanical strength is lower that that of normal tissue.52 As a consequence, scarred regions of the cornea may be more susceptible to rubbing-related trauma. Stromal scarring may be a consequence of direct-rubbing mechanical trauma to keratocytes. During applanation tonometry, collagen fibrils buckle under the compressive tonometer force due to the weak binding forces of the proteoglycan ground substance.53 During IOP measurement of an eye with normal IOP of 10 mm Hg, as little as 1 g of tonometer probe force is sufficient to achieve applanation of the cornea.54 Indentation of the cornea due to rubbing, especially if severe knuckle rubbing and forces of 4.54 kg/2.54 cm2 are involved,43 appears likely to induce a significant mechanical fibrillar disruption. It is possible that indentation induced by rubbing creates awave-like buckling of fibrils that cycle back and forth across the cornea, according to the transfer of rubbing forces. Human cells can be damaged and become necrotic or apoptotic when subjected to mechanical forces (e.g., compression, extension, torsion, shear).55 Because of their predominantly between-lamellae position,42 keratocytes and their network of contacting processes may be exposed to mechanical trauma by buckling and flexure of fibrils, and any associated shear strain (movement of adjacent lamellae relative to one another), as the cornea undergoes repeating back and forth waves of indentation.


A study of the application of experimental digital forces to human eyes indicated that, for an open eye with normal IOP of 15 mm Hg, light and firm digital forces increased IOP by approximately 100% and 300%, respectively.56 The distal end of an index finger was used to apply the light and firm forces to the sclera, through the adnexal skin, during these measurements. 56 The displacement of intraocular fluid and associated IOP spiking may be greater with knuckle rubbing using strong force. The eyelids contribute compressive forces to the ocular surface. For example, relaxed eye closure was found to spike IOP by 2 to 10 mm Hg, and squeeze blinking produced spikes of 50 mm Hg to 110 mm Hg.57,58 Thus, the combination of eye closure and indentation due to digital forces (such as eye rubbing or wiping) may spike IOP to very high levels.3 For example, an unspecified form of massage of a rabbit eye in vivo resulted in IOP .150 mm Hg, which was more than 10 times the physiological level.59 IOP spikes increase with the displacement of intraocular fluid. Increased hydrostatic pressure in corneal tissue can occur due to the increased distending forces of IOP spikes37 that are associated with rubbing-related IOP spikes and related corneal stress stiffening.56 However, as rubbing force oscillates back and forth over the cornea, the area of indentation at any given moment is squeezed between compressive rubbing forces and the distending IOP forces. This combination raises hydrostatic tissue pressure to higher levels than those that are induced by IOP spikes alone. The highest level of hydrostatic tissue pressure is likely to occur at the corneal apex, which is exposed to the highest level of compressive force. A greater degree of corneal indentation— and thus also a higher level of intraocular pressure spiking—may also be due to reduced corneal thickness at the apex. Cells within the intraocular environment are exposed to constant changes in the levels of IOP and associated variations in tissue hydrostatic pressure.37 In vitro studies of cell lines derived from ocular tissues indicated that exposure to sustained levels of elevated hydrostatic pressure (50 mm Hg for periods of up to 6 hours) increases the level of enzyme activity. In addition, there is reorganization of the cell cytoskeleton and associated changes in cell shape.37 Mechanical forces are thought to induce cellular responses through activation of signaling pathways.37 These findings suggest that key intracellular signaling cascades in ocular cells may be similarly affected by elevated IOP in vivo.37 The functional properties of astrocytes can be altered after extremely brief (that is, seconds of) exposure to elevated hydrostatic pressure.60 Functional cellular changes in response to brief elevations in external pressure, combined with the recovery that occurs following cessation of elevated pressure in vitro, may result in significant changes in the homeostatic functions of several intraocular tissues.37 For example, adenylyl cyclase activity was found to be increased in ocular cells after 1 hour of elevated hydrostatic pressure, with recovery to normal levels taking 3 hours.37 The finding that the anterior stromal response to gel swelling is restricted by stromal interweave indicates that a rise in anterior stromal hydrostatic pressure could occur with increased corneal hydration.13 It is possible that keratocytes could be traumatized under conditions of increased hydrostatic pressure and could explain, for example, the preferential loss of anterior stromal keratocytes, which is said to occur in bullous keratopathy.13 Similarly, rubbing-related increased IOP and hydrostatic pressure in the stroma, either alone or in combination with rubbing-related mechanical trauma to the keratocytes, could adversely affect their functions and viability.


One of the mechanisms for the development of KC suggests that the ectasia is due to slippage between collagen fibrils.19 For example, loss of lamellae might be related to a mechanism of rearrangement or sliding of the collagen bundles as the cornea takes a conical shape.19 Alterations in the composition of the ground substance in KC may release the fusion of these bundles, and they may separate through otherwise undetected cleavage patterns.19 It has been suggested20 that the slippage mechanism could be facilitated by enzymatic tissue degradation.16 Mapping collagen orientation by means of x-ray scattering techniques has confirmed that many of the posterior lamellae are displaced, particularly in the region of the cone.20 These observations were found to be consistent with a slippage mechanism that apparently occurs without breaks in the fibrils.20,22,23 Fracture behavior in fiber-reinforced plastic laminate structures appears to provide a basis for understanding mechanisms of loss of corneal strength and development of ectasia. For example, as is the case with fiber-reinforced plastic laminates,61 harsh cyclic rubbing-related loading appears likely to be much more challenging to the cornea than normal levels of distending, compressive, and shear loading due to IOP, blood pressure, and blinking. Similarly, the high potential strength of fibers in plastic laminates is best exploited without exposure to undulating forces.61 Again, compressive stressing that is transverse to the fibers of plastic laminates is potentially very damaging.61 The risk for corneas may be similarly increased with rubbing-related forces that are transverse to fibrils that are susceptible to slippage and perhaps more so for those that have already started slipping from their normal aligned orientation. Delamination and interfiber fracture in fiber-reinforced plastic laminates61 appear to be correlates for interlamellar and interfibrillar slippage that occurs in keratoconus.62 Similarly, fibril fracture in reinforced plastics61 appears to correlate with breaks in Bowman’s layer and ruptures in Descemet’s membrane.62


Surgical dissection of the corneal stroma is not resistance free, even posteriorly, where there is less anteroposterior interweaving.63 This finding suggests that there are elements that bind the collagen together.63 Part of this resistance is due to interactions between the collagen fibrils and other matrix proteins such as the proteoglycans.20 Attachment of keratan sulfate proteoglycans to collagen fibrils was detected in human corneas by means of energy-filtering transmission electron microscopy and atomic force microscopy.64 Large differences between control and KC corneas, in regard to the ordered proteoglycans along the collagen fibrils, have been reported.65 Certain KC proteoglycans were found to take up less stain than their normal control counterparts.65 Cathepsins B and G are known to degrade proteoglycans and collagens, and the finding that these enzymes are upregulated in KC suggests that they may be involved in corneal thinning.17 Three-dimensional studies of the extracellular matrix in human corneal stroma revealed interfibrillar bridging filaments joining neighboring collagen fibrils, like steps of a ladder.21 It is likely that the relatively thin, filamentous cross-bridging structures correspond to the proteoglycan glycosaminoglycan side chains and that the attached globular domains, which bind major collagen fibrils, correspond to the protein cores of proteoglycans.21 Ultrastructural profiles of the extracellular matrix display relatively thick filamentous structures that tightly associate with the surface of major collagen fibrils and project fingerlike structures in the interfibrillar space, to come into contact with the adjacent collagen fibrils.21 Type VI collagen is a major component of the corneal stroma that has been shown to be present in relatively large quantities in human corneas and which is known to interact with proteoglycans.21 The apparent connections of the finger-like structures with major collagen fibrils suggests a bridging or stabilizing function, a role frequently attributed to fibril-associated collagen, with interrupted triple helices, which are known to have the potential to establish other interactions with proteoglycans and type VI collagen.21 If this interfibrillar/interlamellar glue were weakened, the lamellae could tear apart, resulting in displacement of lamellae and redistribution of collagen locally.20 Keratoconus can be explained by an interlamellar and interfibrillar slippage of collagen due to loss of cohesion between cell fibrils and the noncollagen matrix.20 Rubbing-related flexure, and relaxation of lamellae and fibrils in an indented corneal area, may be in contrast to spiked IOP-induced stress stiffening of corneal lamellae in areas adjacent to the indented area. As indentation cycles back and forth over the cornea, the lamellae and fibrils may undergo rapid and violent bulging/indentation alternation between these two conditions, with associated flexure and buckling that could aid slippage.


KC corneas have elevated levels of reactive oxygen species due to an enzyme function imbalance.16,66 An accumulation of reactive oxygen species can greatly damage cells by reacting with proteins, DNA, and membrane phospholipids.16,66 Normally, the natural antioxidant enzymes of the cornea eliminate the reactive oxygen species before they damage cells.16,66 However, oxidative stress can mediate keratocyte apoptosis.16,66 Increased gelatinase-matrix metalloproteinase activities are found in KC corneas, and considerable degradation of the extracellular matrix occurs.16,17 Bovine intramuscular connective tissue was exposed to pressure of 100 to 500 MPa at 8C for 5 minutes.67 High hydrostatic pressure tenderized the samples with structural weakening evident from histological examination and rheometer assessment.67 As described above, IOP of 50 mm Hg alone, for periods up to 6 hours in vitro, can induce human ocular cell responses to the associated increase in hydrostatic pressure. In vivo exposure to the combination of IOP spikes, which may be .50 mm Hg, and associated compressive forces that cyclically squeeze the cornea under higher hydrostatic pressure may induce cellular and/or extracellular responses over short times of exposure. This combination of conditions suggests the possibility that high hydrostatic pressure may initiate structural weakening of corneal tissue during eye rubbing. Compared with the studies of bovine meat samples under conditions of 8C, tenderization  of corneal tissue may be facilitated by a higher temperature (before rubbing) of 34C and possibly a temperature spike during rubbing. In the preparation of meat for consumption, tenderization can be achieved through direct mechanical disturbance (pounding) as well as through slow cooking with moist heat and the action of enzymes.68 Rubbing-related direct mechanical disturbance of keratocytes and/or corneal fibrils due to buckling and flexure associated with indentation may contribute to any other hydrostatic pressure, enzyme, or heat-related processes. This combination may also tend to reduce the rigidity of or tenderize the cornea. In addition, there is the possibility of disturbance of enzyme function due to rubbing trauma,16,66 which could be consistent with the upregulation of adenylyl cyclase in ocular tissue cells under conditions of higher hydrostatic pressure.37 In mammals, matrix metalloproteinases play an essential role in degrading extracellular components.69 This process is analogous with findings that suggest that matrix metalloproteinases are involved in the breakdown of the extracellular matrix and the tenderization of fish muscle tissue.69


The ground substance in the human body is thixotropic.70 Thixotropic (time-dependent) reduction in the viscosity of a gel or other form of pseudoplastic fluid occurs in response to agitation, shaking, squeezing, or other source of shear stress.71 Accordingly, the agitation and shear forces associated with rubbing may reduce the viscosity of the corneal proteoglycan ground substance. Thixotropic loss of viscosity is an active agitation-driven process, whereas recovery, on removal of the influence of agitation, is a passive process. In some cases, the time scales involved can range from minutes, in the case of thixotropic breakdown of viscosity, to hours in rebuilding.71 One of the Weissenberg effects in structured liquids describes the tendency of viscoelastic solutions to flow at right angles to an applied force.72 Consequently, back and forth transfer of rubbing forces across the corneal apex might tend to displace the viscoelastic proteoglycan gel ground substance from the apex by squeezing it toward the peripheral cornea. Reduced viscosity may increase ground substance susceptibility to this form of displacement away from the corneal apex, by cycles of back and forth rubbing forces. Passive recovery of normal apex levels of ground substance viscosity would be expected to occur when rubbing ceases. However, any displacement of ground substance away from the apex may be limited by the capacity for the paracentral and peripheral cornea to accommodate an increased volume of ground substance. However, the apex may be more susceptible to cone-forming rubbing-related IOP spikes, if there is a displacement of ground substance, and associated thinning and weakening (tenderizing) of that region during a rubbing episode. Thixotropic reduction in ground substance viscosity and associated glue function may contribute to lamellar and/or fibrillar slippage in the region of the cone.


The human cornea is known to be strongest to longitudinal stresses (x- and y-axis) and weakest to shear or radial stress (z-axis).22 However, although the normal cornea applanates readily in response to low levels of compressive force during tonometry, it shows a remarkable ability to stressstiffen in response to spiked IOP and the associated increased distending stress.56 Corneal tissue behaves viscoelastically.54 Depending on the nature of mechanical forces to which it is subjected, there may be an (almost) instantaneous deformation, which is the elastic part of the viscoelasticity.73,74 Responses to changes in systolic blood pressure appear to fall into this category. However, viscosity imparts an ability to change shape gradually under stress, with a slow flow, or viscous creep, the speed of which decays with time.73,74 Thus, creep is a time-dependent elongation, and it usually occurs in three stages:73,74 primary creep and decelerating strain rate, secondary creep and constant strain rate, and tertiary creep and accelerating strain rate, which culminates in rupture.73,74 Alternatively, when mechanical stress is removed from a viscoelastic material, there is an instantaneous recovery of the elastic deformation, followed by slow recovery of the viscous creep.73,74 Nevertheless, a viscoelastic material may behave like a liquid, with incomplete recovery and a permanent deformation.73,74 Responses in this category might involve abnormal rubbing with longer duration, and/or greater force, and/or greater frequency. Rubbing-related responses may involve permanent deformation and ectasia, or rupture of Descemet’s membrane and hydrops. In the absence of a history of rubbing, hydrops appears to be a tertiary creep response of a thin cornea to distending IOP. It is possible that an activity that induces an IOP spike that is unrelated to rubbing75 contributes to spontaneous hydrops. Measurements of the cohesive tensile strength of human LASIK corneal wounds indicated a weak central and paracentral stromal scar that averages only 2.4% of the normal corneal strength.76 These laboratory findings correlate well with a clinical series, which indicated that LASIK flaps could be lifted without complications during re-treatments, up to 8.4 years after initial surgery.76 A region of reduced interlamellar cohesive strength, and any associated loss of shear strength, may expose the post-LASIK cornea to ectasia. This risk might be increased by the combination of IOP spikes and tissue-weakening responses to rubbing trauma if a CHAR is resumed or initiated during the postoperative period. Rubbing that is not significant for the normal pre- LASIK cornea may be traumatic for the post-LASIK cornea.3 Studies of interlamellar adhesive strength in human eyebank corneas found that peripheral adhesion is aided by increasing fibril interlamellar weaving.77 Regions of reduced interlamellar cohesive strength vary in magnitude, extent and location between corneas, but with a pronounced drop in the inferior paracentral region.78 The topographic deformation in KC appears to share similar regional specificity with the inherent cohesive strength weakness found in this study.78 Some corneas show a broad inferior weakness that might be associated with an oval cone, and others show a central weakness that might be associated with a round or nipple cone.78 In these regions, adhesion/cohesion may be primarily dependent on the molecular binding strength of proteoglycans.77 Cone formation requires that an area of the cornea is bent from its normal curvature. Bending of the cornea is facilitated by a reduction in shear modulus.79 Plywood consists of an odd number of veneer sheets glued together, with the grains of the veneer layers usually being at right angles to one another.80 The resistance to bending of the relatively flimsy individual veneer sheets is greatly enhanced when they are glued together, so that they are able to resist shear stress and not slide significantly relative to one another. However, in furniture manufacture, steam heating of plywood allows it to be bent to a curved shape. The moist heat may allow the veneer to stretch, with the layer at the convex surface of the curved form requiring an increased arc length, relative to the layer at the concave surface. In addition, heated plywood glue softens, allowing increased shear strain and sliding of the veneer sheets, relative to one another. By this means, bending to the desired curvature can be achieved. When the glue cools, shear strain is again restricted, and the plywood returns to a much greater resistance to additional bending, so that the new curved form is retained. Plywood appears to be a reasonable model for the cornea, which, having preferred inferior/superior and nasal/temporal collagen fibril orientations over the central region of the cornea,81 has a similar right-angled cross-thatching of lamellae. Ordinarily, ground substance might act as a glue to resist shear strain (sliding) responses between lamellae. However, as mentioned above, resistance to corneal bending (and bulging) may be reduced by rubbing agitation and thixotropically reduced viscosity (softening) of the proteoglycan gel ground substance. The resistance to corneal bending also may be reduced if theextracellular matrix is tenderized by any increases in temperature or by any mechanical disturbance associated with indentation, buckling, and flexure.

FIGURE 1. A, an oversize cut of linoleum is clamped by the walls of a room and does not settle to the floor. B, a weight placed centrally results in curvature transfer away from the center, with steepening in the adjacent areas. C, the cornea, clamped at the limbus. D, the cornea, indented by compressive knuckle force. Curvature is transferred from the center of the cornea. Intraocular pressure is raised due to displacement of intraocular fluid and combines with curvature transfer to create stronger cone-forming force.

FIGURE 2. A, a paracentral corneal area that is thinner and/or less rigid. B, central rubbing indentation (RI), high IOP, and curvature transfer combine to increase curvature at x and y. The thinner and/or less rigid area at y may be more likely to yield to this combination of cone-forming mechanisms. The combination of rubbing force and intraocular pressure at the point of indentation (z) creates the highest hydrostatic tissue pressure.


In keratoconus, the cornea becomes weakened and unable to support the intraocular pressure (IOP), the effect of which causes the cornea to bulge forward in the shape of a conical protrusion.10,12,26,27 With IOP approximately doubled, changes in corneal topography were assessed for samples of KC subjects and normal controls by means of an ophthalmodynamometer force to indent the sclera.82 With IOP spiking, stress stiffening in the normal corneas resulted in no significant change at the point of steepest curvature.82 However, for the KC corneas, the IOP spike caused a mean increase at the steepest point of curvature, of 1.84 D (P , 0.018).82 The period of exposure to the IOP spike was only 15 to 20 seconds.81 The change in curvature was not evident when IOP returned to normal levels.82 However, IOP spikes appear to be more likely to contribute to cone formation and/or progression if spiking episodes are longer than 15 to 20 seconds, if they occur frequently, if they involve IOP that is more than double normal levels, and/or if they occur chronically over many years.3,82 Classically, IOP spikes in KC are rubbing-related. However, forms of IOP spiking that are unrelated to rubbing may contribute to cone development and/or progression of KC in subjects with and without a history of a CHAR.75,82 For example, yoga-related inverted body positions, activities involving a Valsalver maneuver, strenuous muscular effort, and lid contact with bedding during prone sleeping can significantly spike IOP.75 Cone formation responses to IOP appear to depend on corneal thinning or other changes to the biomechanical properties of the cornea, which are, for example, due to tenderization and/or reduction in shear modulus. These changes may be genetically based or rubbingrelated, for example. However, rubbing an itchy eye or wiping a watering eye are prime examples of IOP-spiking triggers because they can be very common for some individuals, in 612 | addition to having the potential to be associated with very large IOP spikes. As rubbing forces are transferred back and forth across the cornea, raised IOP increases the distending force on the cornea. Indentation (and associated flattening) of one part of the cornea results in a transfer of curvature and steepening or bulging of another part.26,27,54 This mechanism for steepening of the cornea requires transfer of curvature from a diametrically opposed flattened region.26,27 The inextensibility of fibrils83 and their clamping at the limbus84 appear to be the basis for curvature transfer. The phenomenon is supported by videokeratographic findings in KC.24,25 In those cases, it was found that corneal area was relatively stable during cone progression.24,25 This type of cone formation appears to be more consistent with curvature transfer rather than a slipping or stretching of lamellae to form a cone,24,25 although both mechanisms may be involved. In cases of curvature transfer, steepening associated with inferior cone formation appears to be associated with a pulled curvature transfer from the flattening superior cornea.26,27 Curvature transfer has also been demonstrated in normal corneas fitted with rigid contact lenses.85 In these cases, the superior locating lens flattens the superior cornea, resulting in a pushed curvature transfer to the inferior cornea, where steepening and pseudo-cone formation occur.26,27 Curvature transfer is not dependent on IOP distending forces (Fig. 1), which are, of course, equally distributed over the cornea. However, areas that are adjacent to a rubbing-induced indentation are exposed to both the associated increased distending force of spiked IOP and the associated bulging due to the transfer of curvature from the indented area. Areas that are thinner (the central zone of the normal cornea), or otherwise less resistant to IOP forces (the abnormally thin and/or weakened cone apex in KC), appear to be at greatest risk of a bulging response. The risk may be greatest at those moments when rubbing indentation occurs in an area of the cornea that is adjacent to a thinned and/or weakened region. Under these conditions, the rubbinginduced IOP spike combines with the curvature transfer steepening to increase the chance of cone formation or progression (Fig. 2). The greatest long-term risk appears to be bulging responses that do not fully disappear after a rubbing episode, or those that have not fully dissipated before the next rubbing episode. Failure to return to prerubbing shape may result in the compounding of effects of sequential episodes of mechanical trauma and may facilitate deformation that results in ectasia or ectasia progression.


Corneal scarring prior to contact lens wear, which occurs in association with a KC diagnosis and a history of a CHAR, may be explained by a rubbing-trauma-related wound healing response. For some forms of KC and post-LASIK ectasia, the reduction in shear strength in the ground substance that allows corneal bending and cone formation may be a consequence of rubbing trauma. This process may involve signaling pathway activation, by transduction of rubbing-related mechanical epithelial trauma, that triggers the release of inflammatory mediators and a wound healing response in keratocytes. However, a wound healing response in keratocytes may also be induced by direct rubbing-related mechanical trauma to the keratocytes and/or the effects of increased hydrostatic pressure. Apoptosis rates for keratocytes may also be increased by these mechanisms, and collagen maintenance may be reduced. Ground substance maintenance and its contribution to shear strength may be affected by alterations to keratocytes, but the ground substance may also be altered by abnormal enzyme activity and/or thixotropic reduction in viscosity. Any rubbingrelated increase in corneal temperature may upregulate these mechanisms. Rubbing that occurs before responses to a previous rubbing episode have dissipated may compound some adverse processes. Rubbing-related increased hydrostatic pressure may combine with abnormal enzyme activity to weaken or tenderize the corneal stroma. Large IOP spikes, due to rubbing-related corneal indention, may expose the thinnest/ weakest regions of the cornea to cone-forming bulging. Increased distending forces may induce cone formation by curvature transfer of fibrillar length from a diametrically opposed region of the cornea. Rubbing-related buckling and flexure of fibrils may facilitate cone formation, which is associated with fibrillar slippage in the cone region. Sources of IOP spiking that are unrelated to rubbing may promote cone formation. A cornea that is genetically predisposed to ectasia may be more susceptible to rubbing trauma. The combination of CHAR and a genetic basis to develop KC may increase the potential for rubbing-related trauma and explain early onset, rapid progress, and/or severe forms of KC. The risk of rubbing-related trauma may also be increased by corneal exposure to environmental factors such as ultraviolet radiation and/or contact lens wear.15,16 Regions of reduced strength due to scarring may render those regions more susceptible to rubbing-related trauma. Apart from the possibility of rubbingrelated post-LASIK keratectasia, rubbing may have relevance to other forms of corneal thinning disease or complications following corneal surgery. For example, rubbing-related trauma may have contributed to recurrence of keratoconus in eyes that have been grafted86; to apparently spontaneous bilateral corneal perforation of acute hydrops in KC87; to wound dehiscence in a patient with KC after penetrating keratoplasty and LASIK88; or to globe rupture following keratoplasty in patients with diseases other than keratoconus.89 Prevention is desirable, and there appears to be an indication for the prophylactic control of CHAR.90,91 Some forms of KC may be preventable. Controlling a CHAR after cone development has commenced might slow or otherwise limit the extent of cone development. Rubbing avoidance may reduce the incidence of adverse responses following any form of corneal surgery.


The constructive comments and suggestions contributed by the reviewers of this report are gratefully acknowledged.


1. Krachmer JH, Feder RS, Belin MW. Keratoconus and related noninflammatory corneal thinning disorders. Surv Ophthalmol. 1984;28: 293–322.

2. Rabinowitz YS. Keratoconus. SurvOphthalmol. 1998;42:297–319.

3. McMonnies CW. Abnormal rubbing and keratectasia. Eye Cont Lens. 2007;33:265–271.

4. Bawazeer AM, Hodge WG, Lorimer B. Atopy and keratoconus: a multivariate analysis. Br J Ophthalmol. 2000;84:834–836.

5. Zadnik K, Barr JT, Edrington TB, et al. Baseline findings in the collaborative longitudinal evaluation of keratoconus (CLEK) study. Invest Ophthalmol Vis Sci. 1998;39:2537–2546.

6. Weed KH, MacEwen CJ, Giles T, et al. The Dundee University Scottish Keratoconus study: demographics, corneal signs, associated diseases, and eye rubbing. Eye. 2008;22:534–541.

7. Edwards M, McGhee CNJ, Dean S. The genetics of keratoconus. Clin Exp Ophthalmol. 2001;29:345–351.

8. Sharma, S. Levels of evidence. Evidence-based eye care. 2001;2:5–6.

9. McMonnies CW. The evidentiary significance of case reports: eye rubbing and keratoconus. Optom Vis Sci. 2008;85:262–269.

10. Clifford Hall KG. A comprehensive study of keratoconus. Br J Physiolog Opt. 1963;215–223.

11. Pouliquen Y. Doyne lecture: keratoconus. Eye. 1987;1:1–14.

12. Bron AJ. Keratoconus. Cornea. 1988;7:163–169.

13. Bron AJ. The architecture of the corneal stroma. Br J Ophthalmol. 2001; 85:379–380.

14. Teng CC. Electron microscopic study of the pathology of keratoconus. Part 1. Am J Ophthalmol. 1963;55:18–47.

15. Wilson SE, He Y-G, Qian JW, et al. Epithelial injury induces keratocyte apoptosis: hypothesized role for the interleukin-1 system in the modulation of corneal tissue organization and wound healing. Exp Eye Res. 1996; 62:325–337.

16. Kenney MC, Brown DJ. The cascade hypothesis of keratoconus. Cont Lens Ant Eye. 2003;26:139–146.

17. Zhou L, Sawaguchi S, Twining SS, et al. Expression of degradative enzymes and protease inhibitors in corneas with keratoconus. Invest Ophthalmol Vis Sci. 1998;39:1117–1124.

18. Patey A, Savoldelli M, Pouliquen Y. Keratoconus and normal cornea: a comparative study of the collagenous fibres of the corneal stroma by image analysis. Cornea. 1984;3:119–124.

19. Polack FM. Contributions of electron microscopy to the study of corneal pathology. Surv Ophthalmol. 1976;20:393–395.

20. Meek KM, Tuft SJ, Huang Y, et al. Changes in collagen orientation and distribution in keratoconus corneas. Invest Ophthalmol Vis Sci. 2005;46: 1948–1956.

21. Hirsch M, Prenant G, Renard G. Three-dimensional supramolecular organization of the extracellular matrix in human and rabbit corneal stroma, as revealed by ultrarapid-freezing and deep etching methods. Exp Eye Res. 2001;72:123–135.

22. Dawson DG, Grossniklaus HE, McCarey BE, et al. Biomechanical and wound healing characteristics of corneas after excimer laser keratorefractive surgery: is there a difference between advanced surface ablation and sub-Bowman’s keratomileusis? J Refract Surg. 2008;24:S90–S96.

23. Hayes S, Boote C, Tuft SJ, et al. A study of corneal thickness, shape and collagen organisation in keratoconus using videokeratography and X-ray scattering techniques. Exper Eye Res. 2007;84:423–434.

24. Smolek MK, Klyce SD. Is keratoconus a true ectasia? Arch Ophthalmol. 2000;118:1179–1186.

25. Smolek MK, Klyce SD. Cinemakeratography using computer morphing. Optom Vis Sci. 1997;74:970–975.

26. McMonnies CW. The biomechanics of keratoconus and rigid contact lenses. Eye Contact Lens. 2005;31:80–92.

27. McMonnies CW, SchiefWK. Biomechanically coupled curvature transfer in normal and keratoconus corneal collagen. Eye Cont Lens. 2006;32:51–62.

28. Hill RM.Curiosities of the Contact Lens. Chicago: Professional Press; 1981:66–68.

29. Kessing SV. A new division of the conjunctiva on the basis of X-ray examination. Acta Ophthalmol. 1967;45:680–683.

30. Greiner JV, Peace DG, Baird BS, et al. Effects of eye rubbing on the conjunctiva as a model of ocular inflammation. Am J Ophthalmol. 1985; 100:45–50.

31. Greiner JV, Leahy CD, Welter DA, et al. Histopathology of the ocular surface after eye rubbing . Cornea. 1997;16:327–332.

32. Raizman MB, Rothman JS, Maroun F, et al. Effect of eye rubbing on signs and symptoms of allergic conjunctivitis in cat-sensitive individuals. Ophthalmology. 2000;107:2158–2161.

33. Naoumidi TL, Pallikaris IG, Naoumidi II, et al. Conductive keratoplasty: histological study of human corneas. Am J Ophthalmol. 2005;140:984–992.

34. Fabre EJ, Bureau J, Pouliquen Y, et al. Binding sites for human interleukin 1 alpha, gamma interferon and tumor necrosis factor on cultured fibroblasts of normal cornea and keratoconus. Curr Eye Res. 1991;10:585–592.

35. Kallinikos P, Efon N. On the etiology of keratocyte loss during contact lens wear. Invest Ophthalmol Vis Sci. 2004;45:3011–3020.

36. Kallinikos P, Morgan P, Efron N. Assessment of stromal keratocytes and tear film inflammatory mediators during extended wear of contact lenses. Cornea. 2006;25:1–10.

37. Wax MB, Tezel G, Kobayashi S, et al. Responses of different cell lines from ocular tissues to elevated hydrostatic pressure. Br J Ophthalmol. 2000;84:423–428.

38. Leahy CD, Greiner JV, Hebert E, et al. The effects of eye rubbing on the ocular surface [abstract]. Invest Ophthalmol Vis Sci. 1990;31:208–31.

39. McMonnies CW, Alharbi A, Boneham GC. Epithelial responses to rubbing-related mechanical forces. Cornea. 2009; In publication.

40. Bron AJ. Anterior corneal mosaic. Br J Ophthalmol. 1968;52:659–669.

41. Solomon JD, Case CL, Greiner JV, et al. Warm compress induced visual degradation and Fischer-Schweitzer polygonal reflex. Optom Vis Sci. 2007;84:580–587.

42. Bron AJ, Tripathi RC, Tripathi BJ. Wolff’s Anatomy of the Eye and Orbit. London: Chapman & Hall Medical; 1997:233.

43. Korb DR, Leahy CD, Greiner JV. Prevalence and characteristics of eye rubbing for keratoconic and non-keratoconic subjects. Invest Ophthalmol Vis Sci. 1991;32(Suppl):884.

44. Dursun D, Piniella AM, Pflugfelder SC. Pseudokeratoconus caused by rosacea. Cornea. 2001;20:668–669.

45. Bergmanson JPG. Histopathological analysis of the corneal epithelium after contact lens wear. J Am Optom Assoc. 1987;58:812–818.

46. McMonnies CW, Boneham GC. Keratoconus, allergy, itch, eye-rubbing and hand dominance. Clin Exper Optom. 2003;86:376–384.

47. Baum J. On the location of the cone and the etiology of keratoconus. Cornea. 1995;14:142–143.

48. Iwata JL, Glonek T, Greiner JV, et al Effects of eye rubbing on membrane phospholipids of the cornea [abstract]. Invest Ophthalmol Vis Sci. 1995; 36:3224–3211.

49. Kuo IC. Corneal wound healing. Curr Opin Ophthalmol. 2004;15: 311–315.

50. Zadnik K, Barr JT, Steger-May K, et al. Comparison of flat and steep rigid contact lens fitting methods in keratoconus. Optom Vis Sci. 2005;82: 1014–1021.

51. Wagner H, Barr JT, Zadnik K, et al. Collaborative longitudinal evaluation of keratoconus (CLEK) study: methods and findings to date. Cont Lens Ant Eye. 2007;30:223–232.

52. Matuk Y, McCulloch C. The formation of collagen and its relation to ophthalmic diseases. Can J Ophthalmol. 1980;15:111–116.

53. Kobayashi AS. Mechanics and analysis of tonometry procedures by finite element modeling of the corneoscleral shell. In: Ghista DN, ed. Biomechanics of Medical Devices. New York: Marcel Dekker; 1981: 75–121.

54. Schwartz NJ, Mackay RS, Sackman JL. A theoretical and experimental study of the mechanical behavior of the cornea with application to the measurement of IOP. Bull Mathemat Biophys. 1966;28:585–643.

55. Tan JCH, Kalapesi FB, Coroneo MT. Mechanosensitivity and the eye: cells coping with pressure. Br J Ophthalmol. 2006;90:383–388.

56. McMonnies CW, Boneham GC. Corneal curvature stability with increased intraocular pressure. Eye Cont Lens. 2007;33:124–129.

57. Miller D. Pressure of the lid on the eye. Arch Ophthalmol. 1967;78: 328–330.

58. Coleman DJ, Trokel S. Direct-recorded IOP variations in a human subject. Arch Ophthalmol. 1969;82:637–640.

59. Nash IS, Greene PR, Foster CS. Comparison of mechanical properties of keratoconus and normal corneas. Exp Eye Res. 1982;35:413–423.

60. Islas L, Pasantes-Morales H, Sanchez JA. Characterization of stretchactivated ion channels in cultured astrocytes. Glia. 1993;8:87–96.

61. Knops M. Analysis of Failure in Fiber Polymer Laminates: The Theory of Alfred Puck. Berlin: Springer-Verlag; 2008:5–19.

62. Dawson DG, Ubels JL, Edelhauser HF. Cornea and sclera. In: Kaufman P, Alm A, eds. Adler’s Physiology of the Eye. 11th Edition. St Louis: The C.V. Mosby Company; In publication.

63. Maurice DM. Some puzzles in the microscopic structure of the stroma. J Refract Surg. 1996;12:677–683.

64. Miyagawa A, Kobayashi M, Fujita Y, et al. Surface ultrastructure of collagen fibrils and their association with proteoglycans in human cornea and sclera by atomic force microscopy and energy-filtering transmission electron microscopy. Cornea. 2001;20:651–656.

65. Fullwood NJ, Tuft SJ, Malik NS, et al. Synchroton X-ray diffraction studies of keratoconus corneal stroma. Invest Ophthalmol Vis Sci. 1992; 33:1734–1741.

66. Kenney MC, Chwa M, Atilano SR, et al. Increased levels of catalase and cathepsin V/L2 but decreased TIMP-1 in keratoconus corneas: evidence that oxidative stress plays a role in this disorder. Invest Ophthalmol Vis Sci. 2005;46:823–832.

67. Ichinoseki S, Nishiumi T, Suzuki A. Tenderizing effect of high hydrostatic pressure on bovine intramuscular connective tissue. J Food Sci. 2006;71: E276–E281.

68. Campbell S. The Cook’s Companion. London: Macmillan London Limited; 1980:87.

69. Kubota M, Kinoshita M, Kubota S, et al. Possible implication of metalloproteinases in post-mortem tenderization of fish muscle. Fish Sci. 2001;67:965–968.

70. Hendrickson T. Massage for Orthopedic Conditions. Philadelphia: Lippincott Williams & Wilkins; 2003:9.

71. Barnes HA. Thixotropy: a review. J Non-Newtonian Fluid Mech. 1997;70: 1–33.

72. Collyer AA. Demonstrations with viscoelastic liquids. Phys Educ. 1973;8: 111–116.

73. Vilela PM, Thompson D. Viscoelasticity: why plastic bags give way when you are halfway home. Eur J Phys. 1999;20:15–20.

74. Humphrey JD, Delange SL.An Introduction to Biomechanics: Solids, Fluids, Analysis and Design. New York: Springer; 2004:7–11.

75. McMonnies CW. Intraocular pressure spikes in keratectasia, axial myopia and glaucoma. Optom Vis Sci. 2008;85:1018–1026.

76. Schmack I, Dawson DG, McCarey BE, et al. Cohesive tensile strength of human LASIK wounds with histologic, ultrastructural and clinical correlations. J Ref Surg. 2005;21:433–445.

77. Smolek MK, McCarey BE. Interlamellar adhesive strength in human eyebank corneas. Invest Ophthalmol Vis Sci. 1990;31:1087–1095.

78. Smolek MK. Interlamellar cohesive strength in the vertical meridian of human eye bank corneas. Invest Ophthalmol Vis Sci. 1993;34:2962–2969.

79. Dupps WJ. Biomechanical modeling of corneal ectasia. J Ref Surg. 2005; 21:186–190.

80. Webster’s Collegiate Dictionary. Fifth edition. Springfield: G & C Merriam Co., Publishers, 1947.

81. Meek KM, Newton RH. Organization of collagen fibrils in the corneal stroma in relation to mechanical properties and surgical practice. J Refract Surg. 1999;15:695–699.

82. McMonnies CW, Boneham GC. Corneal responses to intraocular pressure spikes in keratoconus. Cornea. In publication.

83. Berkley DA. Influence of IOP on corneal fluid pressure, tissue stress and thickness. Exp Eye Res. 1971;11:132–139.

84. Doughty MJ, Bergmanson JPG. Collagen fibril characteristics at the corneo-scleral boundary and rabbit corneal swelling. Clin Exper Optom. 2004;87:81–92.

85. Lebow KA, Grohe RM. Differentiating contact lens induced warpage from true keratoconus using corneal topography. CLAO J. 1999;25:114–122.

86. Yeniad B, Alparslan N, Akarcay K. Eye rubbing as an apparent cause of recurrent keratoconus. Cornea. 2009;28:In press.

87. Dantas PE, Nishiwaki-Dantas MC. Spontaneous bilateral corneal perforation of acute hydrops in keratoconus. Eye Cont Lens. 2004;30:40–41.

88. Ranchod TM, McLeod SD. Wound dehiscence in a patient with keratoconus after penetrating keratoplasty and LASIK. Arch Ophthalmol. 2004;122:920–921.

89. Elder MJ, Stack RR. Globe rupture following penetrating keratoplasty: how often, why, and what can we do to prevent it? Cornea. 2004;23: 776–780.

90. McMonnies CW. Management of chronic habits of abnormal rubbing. Cont Lens Ant Eye. 2008;31:95–102.

91. McMonnies CW. Behaviour modification in the management of chronic habits of abnormal eye rubbing. Cont Lens Ant Eye. 2009;35:In press.