[Skip to Navigation]
Sign In
Figure 1.  Examples of Contact Lens Sensor (CLS) Values vs Time During Testing
Examples of Contact Lens Sensor (CLS) Values vs Time During Testing

The CLS data from 2 patients with glaucoma are shown. The blue line represents SENSIMED AG algorithm median values, while the black line represents the mean values. Solid circles and triangles show the corresponding 30-second interval median and mean CLS values. The colored shading indicates patient body position. The intraocular pressure levels near the top of each graph are tonometer data from the start and end of the protocol in sitting and lying positions.

Figure 2.  Change in Contact Lens Sensor (CLS) Output by Position for All Study Participants and by Patient Group
Change in Contact Lens Sensor (CLS) Output by Position for All Study Participants and by Patient Group

The mean CLS values are estimated from generalized estimating equation models for the 5 body positions tested, including lateral decubitus (lying-at-start [LS]), FD1, supine (relax-1 [R1]), FD2, and supine (relax-2 [R2]) for durations indicated in the Methods section. The mean (SE) values are shown. The trend lines were generated by the generalized estimating equation linear regression model when the 5 body positions are sequentially assigned values 1 through 5. FD indicates face down.

Figure 3.  Contact Lens Sensor (CLS) Values During FD1 in Patients With Glaucoma
Contact Lens Sensor (CLS) Values During FD1 in Patients With Glaucoma

The linear regression line slope from the generalized estimating equation model is 0.275 mV Eq/min (P = .53). The values plotted represent change in CLS from the mean CLS lying-at-start (LS) to standardize across participants. FD indicates face down.

Figure 4.  Estimation Nomogram Relating Modeled and Measured Changes in Limbal Strain and in Intraocular Pressure (IOP)
Estimation Nomogram Relating Modeled and Measured Changes in Limbal Strain and in Intraocular Pressure (IOP)

Solid lines show the modeled limbal strain vs change in IOP for known initial values of IOP. The right axis shows contact lens sensor (CLS) values in patients with glaucoma when moving to and from face down (FD) position (the mean ±1 SE values are shown as dashed and dotted orange lines, respectively). The right axis shows that the mean (SE) CLS value of 40.9 (18.0) mV Eq in glaucoma eyes intersects the initial IOP equals 14.2 mm Hg line at the blue diamonds, where the mean (SE) IOP increase is 2.5 (1.1) mm Hg, which in turn corresponds to the mean (SE) modeled strain increase (on the left axis) of 283 (116) microstrain.

Table.  Estimated Mean Change in CLS Value by Position and by Patient Group
Estimated Mean Change in CLS Value by Position and by Patient Group
1.
Quigley  HA, Broman  AT.  The number of people with glaucoma worldwide in 2010 and 2020.  Br J Ophthalmol. 2006;90(3):262-267.PubMedGoogle ScholarCrossref
2.
Boland  MV, Quigley  HA.  Risk factors and open-angle glaucoma.  J Glaucoma. 2007;16(4):406-418.PubMedGoogle ScholarCrossref
3.
Bengtsson  B, Heijl  A.  Diurnal IOP fluctuation: not an independent risk factor for glaucomatous visual field loss in high-risk ocular hypertension.  Graefes Arch Clin Exp Ophthalmol. 2005;243(6):513-518.PubMedGoogle ScholarCrossref
4.
De Moraes  CG, Juthani  VJ, Liebmann  JM,  et al.  Risk factors for visual field progression in treated glaucoma.  Arch Ophthalmol. 2011;129(5):562-568.PubMedGoogle ScholarCrossref
5.
Nouri-Mahdavi  K, Hoffman  D, Coleman  AL,  et al; Advanced Glaucoma Intervention Study.  Predictive factors for glaucomatous visual field progression in the Advanced Glaucoma Intervention Study.  Ophthalmology. 2004;111(9):1627-1635.PubMedGoogle ScholarCrossref
6.
Heijl  A, Leske  MC, Bengtsson  B, Hyman  L, Bengtsson  B, Hussein  M; Early Manifest Glaucoma Trial Group.  Reduction of intraocular pressure and glaucoma progression.  Arch Ophthalmol. 2002;120(10):1268-1279.PubMedGoogle ScholarCrossref
7.
Quigley  HA, Addicks  EM, Green  WR, Maumenee  AE.  Optic nerve damage in human glaucoma, II.  Arch Ophthalmol. 1981;99(4):635-649.PubMedGoogle ScholarCrossref
8.
Quigley  HA, Cone  FE.  Development of diagnostic and treatment strategies for glaucoma through understanding and modification of scleral and lamina cribrosa connective tissue.  Cell Tissue Res. 2013;353(2):231-244.PubMedGoogle ScholarCrossref
9.
Norman  RE, Flanagan  JG, Sigal  IA, Rausch  SM, Tertinegg  I, Ethier  CR.  Finite element modeling of the human sclera.  Exp Eye Res. 2011;93(1):4-12.PubMedGoogle ScholarCrossref
10.
Sigal  IA, Yang  H, Roberts  MD, Burgoyne  CF, Downs  JC.  IOP-induced lamina cribrosa displacement and scleral canal expansion.  Invest Ophthalmol Vis Sci. 2011;52(3):1896-1907.PubMedGoogle ScholarCrossref
11.
Girard  MJ, Suh  JK, Bottlang  M, Burgoyne  CF, Downs  JC.  Scleral biomechanics in the aging monkey eye.  Invest Ophthalmol Vis Sci. 2009;50(11):5226-5237.PubMedGoogle ScholarCrossref
12.
Broman  AT, Quigley  HA, West  SK,  et al.  Estimating the rate of progressive visual field damage in those with open-angle glaucoma, from cross-sectional data.  Invest Ophthalmol Vis Sci. 2008;49(1):66-76.PubMedGoogle ScholarCrossref
13.
Kim  KN, Jeoung  JW, Park  KH, Lee  DS, Kim  DM.  Effect of lateral decubitus position on intraocular pressure in glaucoma patients with asymmetric visual field loss.  Ophthalmology. 2013;120(4):731-735.PubMedGoogle ScholarCrossref
14.
Korenfeld  MS, Dueker  DK.  Occult intraocular pressure elevations and optic cup asymmetry.  Invest Ophthalmol Vis Sci. 1993;34(suppl):. Abstract 1433.Google Scholar
15.
Flatau  AB.  Mitigation of the IOP elevation associated with non-supine sleep positions.  Acta Ophthalmologica.2011;89(suppl 248): Abstract 2256.Google Scholar
16.
Lee  SY, Jeoung  JW, Kim  DM, Park  KH.  Relationship between sleeping position and asymmetric visual-field loss in glaucoma patients by quantitative measurement of body position using a mobile device  [ARVO Abstract 112].  Invest Ophthalmol Vis Sci. 2015;56(7):112.Google Scholar
17.
Kim  KN, Jeoung  JW, Park  KH, Kim  DM, Ritch  R.  Relationship between preferred sleeping position and asymmetric visual field loss in open-angle glaucoma patients.  Am J Ophthalmol. 2014;157(3):739-745.PubMedGoogle ScholarCrossref
18.
Volpe  P, Flatau  A. Models of stress, strain deformation and IOP caused by non-supine sleep-position–induced loads. In: Proceedings of the BMES Frontiers in Medical Devices Conference; May 18-20, 2015; Washington, DC. Track 4, Abstract P15.
19.
Iacangelo  A, Ko  H, Kozlowski  Y, Prendergast  M, Zubajlo  R, Flatau  A. Measuring and modeling facial features for protective eye shield design. In: Proceedings of the BMES Frontiers in Medical Devices Conference; May 18-20, 2015; Washington, DC. Track 3, Abstract P73.
20.
Meurs  I, Thepass  G, Lemij  HG, Stuij  A.  Is a pillow a risk factor for glaucoma?  [ARVO Abstract 2024].  Invest Ophthalmol Vis Sci. 2015;56(7):2024.Google Scholar
21.
Mansouri  K, Medeiros  FA, Tafreshi  A, Weinreb  RN.  Continuous 24-hour monitoring of intraocular pressure patterns with a contact lens sensor.  Arch Ophthalmol. 2012;130(12):1534-1539. Medline:PubMedGoogle ScholarCrossref
22.
Lorenz  K, Korb  C, Herzog  N,  et al.  Tolerability of 24-hour intraocular pressure monitoring of a pressure-sensitive contact lens.  J Glaucoma. 2013;22(4):311-316.PubMedGoogle ScholarCrossref
23.
Mansouri  K.  The road ahead to continuous 24-hour intraocular pressure monitoring in glaucoma.  J Ophthalmic Vis Res. 2014;9(2):260-268.PubMedGoogle Scholar
24.
Mansouri  K, Shaarawy  T.  Continuous intraocular pressure monitoring with a wireless ocular telemetry sensor.  Br J Ophthalmol. 2011;95(5):627-629.PubMedGoogle ScholarCrossref
25.
De Smedt  S, Mermoud  A, Schnyder  C.  24-Hour intraocular pressure fluctuation monitoring using an ocular telemetry Sensor.  J Glaucoma. 2012;21(8):539-544.PubMedGoogle ScholarCrossref
26.
Aptel  F, Tamisier  R, Pépin  JL,  et al.  Hourly awakening vs continuous contact lens sensor measurements of 24-hour intraocular pressure.  JAMA Ophthalmol. 2014;132(10):1232-1238.PubMedGoogle ScholarCrossref
27.
Agnifili  L, Mastropasqua  R, Frezzotti  P,  et al.  Circadian intraocular pressure patterns in healthy subjects, primary open angle and normal tension glaucoma patients with a contact lens sensor.  Acta Ophthalmol. 2015;93(1):e14-e21. doi:10.1111/aos.12408.PubMedGoogle ScholarCrossref
28.
Leonardi  M, Pitchon  EM, Bertsch  A, Renaud  P, Mermoud  A.  Wireless contact lens sensor for intraocular pressure monitoring.  Acta Ophthalmol. 2009;87(4):433-437.PubMedGoogle ScholarCrossref
29.
Mansouri  K, Weinreb  RN, Liu  JH.  Efficacy of a contact lens sensor for monitoring 24-h intraocular pressure related patterns.  PLoS One. 2015;10(5):e0125530. doi:10.1371/journal.pone.0125530.PubMedGoogle ScholarCrossref
30.
World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.  JAMA. 2013;310(20):2191-2194. doi:10.1001/jama.2013.281053.PubMedGoogle ScholarCrossref
31.
Silver  DM, Geyer  O.  Pressure-volume relation for the living human eye.  Curr Eye Res. 2000;20(2):115-120.PubMedGoogle ScholarCrossref
32.
Chandrupatla  TR, Osler  TJ.  The perimeter of an ellipse.  Math Scientist. 2010;35(2):122-131.Google Scholar
33.
Wolfram Demonstrations Project. Bryant  J. Ellipsoid. http://demonstrations.wolfram.com/Ellipsoid/. Accessed July 17, 2015.
34.
Coudrillier  B, Tian  J, Alexander  S, Myers  KM, Quigley  HA, Nguyen  TD.  Biomechanics of the human posterior sclera.  Invest Ophthalmol Vis Sci. 2012;53(4):1714-1728.PubMedGoogle ScholarCrossref
35.
Hommer  A, Fuchsjäger-Mayrl  G, Resch  H, Vass  C, Garhofer  G, Schmetterer  L.  Estimation of ocular rigidity based on measurement of pulse amplitude using pneumotonometry and fundus pulse using laser interferometry in glaucoma.  Invest Ophthalmol Vis Sci. 2008;49(9):4046-4050.PubMedGoogle ScholarCrossref
36.
Downs  JC, Suh  JK, Thomas  KA, Bellezza  AJ, Hart  RT, Burgoyne  CF.  Viscoelastic material properties of the peripapillary sclera in normal and early-glaucoma monkey eyes.  Invest Ophthalmol Vis Sci. 2005;46(2):540-546.PubMedGoogle ScholarCrossref
37.
Nguyen  C, Cone  FE, Nguyen  TD,  et al.  Studies of scleral biomechanical behavior related to susceptibility for retinal ganglion cell loss in experimental mouse glaucoma.  Invest Ophthalmol Vis Sci. 2013;54(3):1767-1780.PubMedGoogle ScholarCrossref
38.
Kimball  EC, Nguyen  C, Steinhart  MR,  et al.  Experimental scleral cross-linking increases glaucoma damage in a mouse model.  Exp Eye Res. 2014;128:129-140.PubMedGoogle ScholarCrossref
39.
Gordon  MO, Beiser  JA, Brandt  JD,  et al.  The Ocular Hypertension Treatment Study.  Arch Ophthalmol. 2002;120(6):714-720.PubMedGoogle ScholarCrossref
Original Investigation
April 2016

Measured Changes in Limbal Strain During Simulated Sleep in Face Down Position Using an Instrumented Contact Lens in Healthy Adults and Adults With Glaucoma

Author Affiliations
  • 1Department of Aerospace Engineering, University of Maryland, College Park
  • 2The Fischell Department of Bioengineering, University of Maryland, College Park
  • 3Glaucoma Center of Excellence, Wilmer Eye Institute, The Johns Hopkins University School of Medicine, Baltimore, Maryland
JAMA Ophthalmol. 2016;134(4):375-382. doi:10.1001/jamaophthalmol.2015.5667
Abstract

Importance  Eyes of patients with glaucoma may be damaged during sleep.

Objective  To measure strains in glaucoma eyes and control eyes produced by mechanical force or deformation of the eye from contact when one side of the face rests against a pillow.

Design, Setting, and Participants  This study took place in a clinic-based setting among 22 patients with glaucoma and 11 age-matched controls. The research was conducted at Wilmer Eye Institute between February 4, 2014, and December 2, 2014. Data analysis was done from June 3, 2014, to June 30, 2015.

Main Outcomes and Measures  We used a contact lens sensor (CLS) to measure change in limbal strain associated with placing one side of the face down (FD) on a pillow in simulated sleep. Baseline intraocular pressure (IOP) was measured with a tonometer. The CLS data were collected every 5 minutes during intervals of up to 60 minutes in various positions, including sitting, lateral decubitus, FD (with the CLS-instrumented eye toward the pillow), and supine. Measured changes in limbal strain were related to estimated changes in IOP and to modeled strain produced by changes in IOP.

Results  Among 22 patients with glaucoma and 11 controls, 17 were female. The mean age for the glaucoma group was 62.6 years, while the mean age for the control group was 61.4 years (P = .68). Baseline IOP was also similar for the 2 groups. The mean IOP sitting at the start was 13.7 mm Hg for the glaucoma group and 13.8 mm Hg for the control group (P = .73), and the mean IOP lying at the start was 17.5 mm Hg for the glaucoma group and 16.0 mm Hg for the control group (P = .88). By multivariable linear regression, FD position was associated with an increase in limbal strain in glaucoma eyes (mean [SE], 44.1 [20.4] mV Eq; P = .03) but not in control eyes (mean [SE], 13.6 [13.9] mV Eq, P = .33). While FD, the increased CLS values in patients with glaucoma did not decrease over time (slope, 0.275 mV Eq/min; P = .53 by univariable linear regression). Magnitudes of measured changes in limbal strain were greater in glaucoma eyes with past visual field worsening (P = .006 by multivariable linear modeling). The mean limbal strain increase among patients with glaucoma in FD position was equivalent to strain expected for a mean (SE) IOP increase of 2.5 (1.1) mm Hg from a baseline IOP of 14.2 mm Hg.

Conclusions and Relevance  Contact with a pillow in FD position during simulated sleep produced a sustained strain increase in glaucoma eyes, particularly those eyes with past progressive visual field loss. The mean FD change in glaucoma eyes was equivalent to strain increase associated with a mean (SE) sustained IOP elevation of 2.5 (1.1) mm Hg. Further experiments are planned to determine if facial features or a protective eye shield prevents sleep position–induced limbal strains during a mean 8-hour sleep period.

Introduction

Glaucoma is the second leading cause of vision loss worldwide.1 Risk factors for open-angle glaucoma2 (OAG) include the mean level of intraocular pressure (IOP)3 and its fluctuation.4,5 Lowering of IOP protects against progressive visual loss.6 Intraocular pressure acts as a mechanical load on the optic nerve head through the translaminar pressure differential and through tensile hoop stresses in the sclera, leading to ganglion cell axon damage.7 Variation in ocular connective tissue mechanical properties may explain why half of the individuals with OAG experience injury at physiological IOP levels.8 Biomechanical models suggest that cornea and sclera mechanical behavior influences optic nerve head deformation.9-11

While both OAG and angle-closure glaucoma have strong genetic components, environmental risk factors may have an important role in their morbidity because asymmetry of glaucoma damage is the rule rather than the exception.12 The concept of sleep position contributing to asymmetric vision loss in patients with glaucoma has been proposed previously,13-16 and some evidence suggests a relationship between self-reported preferred side of sleep and the eye with greater visual field loss.13,14,17 Potential prevention of IOP-induced or deformation-induced strain in sleep positions by protective eye shields has been proposed,15,18-20 but measures of IOP or strain during sleep have not been possible until recently. A contact lens sensor (CLS) (Triggerfish; SENSIMED AG)21-24 incorporates a strain gauge at the limbus that measures strain associated with IOP change or mechanical force on the eye with open or closed eyelids. Its output is transmitted to an antenna taped to the face that is connected to a hardware device. The CLS tracks changes in limbal strain for periods of many hours, including during sleep.25-27 Changes in CLS output have been correlated with changes in IOP when only air (not a pillow) contacts the eye.28,29

We investigated the effect of sleep position on CLS-measured limbal strain in several positions, including a face down (FD) position that is hypothesized to produce sustained deformation of the eye and potentially cause sustained IOP elevation. Herein, we compare strain measurements in patients with glaucoma vs age-matched controls.

Box Section Ref ID

Key Points

  • Question: Does limbal strain differ between control eyes and glaucoma eyes when in simulated face down sleep positions with one side of the face against a pillow?

  • Findings: This case-control study showed a significant increase in limbal strain in patients with glaucoma, not in controls, especially those with past visual field worsening. Mean face down strain increase among participants with glaucoma was equivalent to that expected from an increase in intraocular pressure of 2.5±1.1 mm Hg.

  • Meaning: Face contact with a pillow during simulated sleep produced a sustained strain increase in eyes with glaucoma, particularly those with past progressive visual field loss.

Methods

The research was conducted at Wilmer Eye Institute between February 4, 2014, and December 2, 2014, on one eye each of 22 patients with glaucoma and 11 healthy controls. Data analysis was done from June 3, 2014, to June 30, 2015. Study participants were at least 18 years old. Individuals with glaucoma were patients at the Wilmer Eye Institute. Controls were volunteers or unrelated persons accompanying the patients. Control eyes were examined to assure nondisease status. The study was approved and monitored by the institutional review board of The Johns Hopkins University School of Medicine, written informed consent was obtained for each patient, and the work abided by the Declaration of Helsinki.30

We excluded individuals with past intolerance to contact lenses and individuals who had undergone retinal reattachment surgery, trabeculectomy, tube shunt, or diode glaucoma surgical procedures. We also excluded individuals who had allergies to tape, bandage adhesives, or gel.

Before testing, demographic and clinical data (eTable 1 in the Supplement) were gathered. Eleven patients with glaucoma who had undergone 5 or more past visual field tests were categorized into one of 3 groups, including past progressive visual field worsening, stable fields, or indeterminate. Past visual field testing was reviewed by a glaucoma specialist (H.A.Q.) who was masked to the IOP data. Patients were categorized based on the mean deviation, pattern standard deviation, and glaucoma progression analysis.

A tonometer (Icare TA01i; Icare Finland Oy) was used to measure IOP in the study eye when in sitting and lateral decubitus positions. Six measures at low variance were collected 3 times for each position, and the mean IOP was calculated for each position. The study eye was inferior for lateral decubitus and FD positions (eFigure 1 in the Supplement).

The CLS was matched to each individual’s central corneal radius and placed on the eye. Individuals moved through the following sequence of positions: supine for 60 minutes, sitting for 7.5 minutes (sitting-at-start), lateral decubitus for 12 minutes (lying-at-start), FD for 60 minutes (FD1), supine for 15 minutes (relax-1), FD for 30 minutes (FD2), supine for 10 minutes (relax-2), lateral decubitus for 7.5 minutes, and sitting for 10 minutes. The CLS was removed, and the individual rested in a supine position for 20 minutes. The tonometer was then used to measure IOP in lateral decubitus and sitting positions. All testing was performed during daytime hours. Individuals wore comfortable clothing without a tight neck or necktie. They were asked to avoid large volumes of fluid intake for 1 hour before the study.

The CLS data were acquired over approximately 3 hours. The CLS data logger wirelessly interrogates the CLS for 30 seconds at 10 Hz every 5 minutes. A log of study participant movements was recorded, with variations to the intended protocol noted. The CLS output in units of millivolt equivalents is initially set to zero by system software. Magnitude of change in limbal strain with position was the primary outcome variable.

Raw CLS data are reported using the SENSIMED AG algorithm to calculate the medians of the 300 data points gathered during each 30-second sampling interval. The algorithm mitigates effects of blinking and other disturbances and provides values that are similar to the mathematical mean of the raw data (Figure 1).

We modeled changes in limbal strain due to changes in IOP using the model by Silver and Geyer31 for the living human eye. We used their approach of modeling the eye as an ellipsoid and the pressure-volume relationship in their equation 7 to calculate changes in volume and semiprincipal axis lengths associated with rises in IOP (ΔP) for different initial IOP values (P0). The equatorial circumference lengths L(P) as a function of P0 and ΔP were calculated using the Gauss-Kummer method for determining the perimeter of an ellipse.32 Dimensions for semimajor and semiminor axes of the 2-dimensional elliptical cross section at the equatorial plane of the eye were taken from Table 2 in the article by Silver and Geyer.31 We also approximated this cross section as a circle (diameter D as the mean of semimajor and semiminor axes). Lengths L(P0P) and L(P0) from the elliptical and circular models were used to determine equatorial strain in the following equation:

The models give similar results, with difference in length L(P0P) less than 0.002 mm and in ε(P0P)equatorial less than 0.001%. In addition, noting that cross sections perpendicular to the anterior-posterior axis of an ellipsoid (sphere) are similar ellipses (circles),33 fractional change in cross-sectional diameter (Δ) is used in the following equation to show that, for any cross-sectional diameter and initial pressure P0, the ΔP-induced circumferential strain is equal to Δ:

For statistical analysis, demographic and clinical data and baseline tonometer IOP were summarized separately for patients with glaucoma and controls. For each variable, the 2 groups were compared using the Fisher exact test or the Freeman-Halton exact test for categorical variables and the Wilcoxon rank sum test for continuous variables. The tonometer IOP and the mean CLS value obtained with each participant sitting-at-start and lying-at-start were compared. Spearman rank correlation coefficients were used because some of the measures did not appear to be normally distributed.

All estimates and P values for change in CLS with position are derived from generalized estimating equation (GEE) linear models that take into account correlations among repeated measurements on a given participant. After examination of the Akaike information criterion for several alternatives, the working correlation matrix for the repeated CLS measurements in a given position and participant was assumed to have a first-order autoregressive structure, in which measurements taken closer in time have higher correlation. The dependent variable for these models was the CLS value. Independent variables were position, group, or a demographic or clinical characteristic, as well as additional terms for interactions if appropriate. The Bonferroni method was used to adjust pairwise significance levels for multiple comparisons.

To assess the linear trend in CLS measures during FD1, the independent variable for the GEE model was the FD1 CLS value minus the mean CLS for the previous position to standardize across participants. The independent variable for this model was the number of minutes the participant had been in the FD1 position. A local regression curve was also used to assess the trend over time. All statistical analyses were performed using a software program (SAS, version 9.2; SAS Institute).

Results

Differences between the glaucoma and control groups are summarized in eTable 2 in the Supplement, with the only notable difference being the cup-disc ratio, as expected (P < .001). All but one patient with glaucoma had OAG, and the single exception had angle-closure glaucoma. Most were using 1 or more types of glaucoma eye drops daily.

Before CLS insertion, the mean (SE) starting IOP values measured with the tonometer in sitting and lying positions for the 11 controls were 13.8 (1.1) mm Hg and 16.0 (1.2) mm Hg, respectively. For the 22 patients with glaucoma, the mean (SE) starting IOP values for sitting and lying were 13.7 (1.1) mm Hg and 17.5 (1.4) mm Hg, respectively (P = .73 comparing groups for sitting and P = .88 comparing groups for lying). For both groups, an increase in IOP from sitting to lying positions was observed in the tonometer data (P = .004 for control eyes and P < .001 for glaucoma eyes, Wilcoxon signed-rank test). An increase from sitting to lying in CLS values was not evident (P = .96 for control eyes and P = .61 for glaucoma eyes, GEE model). Comparison between tonometer and CLS values at comparable sitting-at-start and lying-at-start body positions had substantial variability, as shown in eFigure 2 in the Supplement.

Each patient data set consisted of several thousand raw CLS values (approximately 44 intervals 30 seconds long of approximately 300 data points each), which were linked to position during testing (Figure 1). Data for all patients were entered into GEE linear regression models to estimate differences in the mean CLS value. Overall, the 33 participants showed an increase in the mean CLS value when moving to and from FD positions (P = .004) (Figure 2A), where higher CLS value was correlated with higher tonometric IOP (eFigure 2 in the Supplement). Individual positional changes in the mean (SE) CLS units (±1 SE) included lying-at-start to FD1 (36.1 [14.5], P = .01), FD1 to relax-1 (33.6 [17.5], P = .06), relax-1 to FD2 (34.8 [10.6], P = .001), and FD2 to relax-2 (23.1 [10.4], P = .03) (Table). Adjusting for multiple comparisons, the P value for the increase in CLS values from relax-1 to FD2 was P = .006. The mean (SE) FD1 and FD2 difference was 1.3 (15.1) CLS units (P = .93), indicating a consistent increase in CLS values with each shift to the FD position.

The patients with glaucoma showed an increase in their mean CLS value when moving to and from FD positions (Figure 2B), while the controls did not (P = .003 and P = .73, respectively) (Table). For patients with glaucoma, the mean (SE) increases in CLS units were 44.1 (20.4) (P = .03) for lying-at-start to FD1, 39.5 (23.7) (P = .10) for FD1 to relax-1, 47.9 (13.8) (P = .001) for relax-1 to FD2, and 32.1 (14.0) (P = .02) for FD2 to relax-2. The P value for the increase from relax-1 to FD2 was P = .003 after adjusting for multiple comparisons.

Comparing all individual features with the change in CLS values on moving to the FD positions, we found that smaller cup-disc ratio was associated with greater increase in CLS value among patients with glaucoma (P = .001 overall). This finding seemed dependent on substantial change in 2 eyes with small cup-disc ratios, while most of the glaucoma eyes had cup size of 0.7 or greater.

Of the 22 glaucoma eyes, 12 had 5 or more fields for assessment of progression. Eight of 11 were judged progressive in one or both eyes, and 5 of 11 were progressive in the CLS-tested eye (the 12th was categorized as indeterminate in progression). The change in CLS value moving to FD was greater in the glaucoma eyes that were judged in masked evaluation to have past progressive visual field loss in one or both eyes (P = .006 overall). The P value for comparison between FD1 and lying-at-start was P = .05 after applying the Bonferroni correction. The CLS value increase was also greater in glaucoma eyes with past progressive visual field loss in the tested eye alone (P = .05 overall).

None of the other features listed in eTable 1 in the Supplement were statistically related to the CLS value changes, possibly because of the limited study size. This finding includes subjective ratings of facial features and structure, which we are reassessing with newer quantitative tools.

There was diversity in the change in CLS values with positional change among the glaucoma group, with some having large and sustained increases, some demonstrating increases that diminished with time, and some exhibiting minimal strain change (Figure 1). For patients with glaucoma, the estimated change in CLS units over time was 0.275 mV Eq/min (P = .53) during FD1, indicating no trend toward diminution of the rise in CLS units over time (Figure 3).

The increase in tonometer IOP sitting-to-lying at the end of the study was equated to the rise in the mean CLS values sitting-at-end to lying-at-end (ie, just before removing the CLS). The mean (SE) rise in tonometer IOP sitting-to-lying across all 33 participants was 3.2 (0.4) mm Hg (P < .001, GEE model) from a sitting mean of 14.2 mm Hg. The corresponding mean (SE) rise in CLS units was 51.2 (13.1) mV Eq (P < .001, GEE model). Comparison of the means for these similar (in-air) conditions provides calibration of our CLS strain data to a known change in IOP from a mean initial IOP of 14.2 mm Hg.

Discussion

To our knowledge, the findings herein demonstrate for the first time that moving to FD sleep positions leads to an increase in ocular strain that is sustained when the individual remains in the FD position. Glaucoma eyes, particularly in patients with past visual field progression, had an increase in CLS values on FD positioning not seen in control eyes. The eyes of patients with glaucoma experienced a mean increase in limbal strain that was 3 times greater than that of the eyes of controls in the FD position. In fact, the glaucoma eyes constituted almost all of those eyes that had increases of a measurable degree. The greater deformation in glaucoma eyes, particularly those eyes with progressive visual field loss, than in control eyes could derive from the features of the facial configuration of the individual or from the response of the cornea and sclera to stress. We are investigating how change in CLS strain varies with facial features and structure using 3-dimensional scans and studying the manner in which the orbit is applied to the pillow as each person sleeps.

If the increase in CLS value is related to the stress-strain response of the globe itself, this finding suggests that the limbal area of glaucoma eyes that we tested had greater strain (more compliance) than control eyes. Glaucoma eyes could be different from nonglaucoma eyes at baseline, or they might become more compliant after remodeling induced by the disease or its treatment. We do not have longitudinal data that would allow measurement of change over time in the same individual to distinguish among these possible explanations. There is evidence that the posterior, peripapillary sclera of glaucoma eyes is stiffer than that of age-matched controls in inflation testing34 and by indirect in vivo methods.35 In experimental monkey36 and mouse37 glaucoma, chronic elevated IOP leads to greater posterior scleral stiffness. An experimental study38 in mice found that induction of greater scleral stiffness led to greater retinal ganglion cell loss. However, it is feasible that the anterior-posterior ocular structures are not matched in their mechanical behavior. Furthermore, all but one of the patients with glaucoma herein had been treated with eyedrop medication, and more than half of them were taking prostaglandin analogues, which could affect anterior mechanical responses.

The mean level of CLS values during the FD positioning did not trend back toward lower values during the 60-minute period during which the face was in contact with the pillow (Figure 3). Further experiments are planned to determine if elevated strains are sustained during a mean 8-hour sleep period and if a protective eye shield prevents sleep position–induced mechanical force and deformation from producing limbal strain. The higher strain in the glaucoma eyes during FD positioning returns toward baseline after FD positioning but elevates again on FD positioning, leading to greater fluctuation in strain during sleep. Higher fluctuation in IOP is thought by some to be an independent risk factor for glaucoma damage,5 adding to higher mean IOP. In addition, a positive trend line slope was observed for patients with glaucoma but not controls (Figure 2B).

Our GEE regression analysis predicts that higher initial IOP values will have smaller increases in CLS strain on moving to the FD position. Using Figure 4 to assess the associated change in IOP for the predicted change in strain for selected initial IOPs, we found that higher initial IOP values have smaller increases in IOP.

The mean (SE) FD change in limbal strain for the 22 patients with glaucoma moving to and from FD was 40.9 (18.0) mV Eq. This line is superimposed on the modeled strain response curves in Figure 4 using the in-air calibration for CLS units to IOP (16.2 mV Eq/mm Hg sitting-to-lying from the sitting mean IOP of 14.2 mm Hg). The CLS mean (SE) strain intersects the 14.2 mm Hg trace at an IOP increase of 2.5 (1.1) mm Hg, which in turn corresponds to the mean (SE) modeled strain increase of 283 (116) microstrain. The FD strains are similar in magnitude to strains produced by IOP increases of approximately 1 to 5 mm Hg for the modeled initial IOP values (10-25 mm Hg).

Intraocular pressure differences of a small magnitude over extended periods may be important contributors to glaucoma damage. The Early Manifest Glaucoma Trial6 and the Ocular Hypertension Treatment Study39 found a 10% increase in risk (improvement) for each increase (reduction) of 1 mm Hg in the mean IOP.

The present research is subject to limitations. The CLS device produces an output value that is not directly linked to IOP. The model assumes that eccentricity at the limbus is small, and measured strain is insensitive to small variations in CLS placement. In asking each participant to assume the FD position, we were perhaps not simulating that individual’s normal sleep position. The study did not obtain measurements during actual sleep, which could affect the outcomes in ways we cannot predict. There were too few participants using each type of medication to test for effects of individual classes of glaucoma drugs.

Conclusions

Moving to the FD sleep position produced increased mean strain in the eyes of patients with glaucoma. These data point to the need for studies of preventive strategies among at least some persons with glaucoma. Further study is needed to identify attributes of patients with glaucoma who are likely to benefit from preventing FD-induced strain and deformation of the eye.

Back to top
Article Information

Submitted for Publication: July 22, 2015; final revision received November 22, 2015; accepted November 23, 2015.

Corresponding Author: Alison Flatau, PhD, Department of Aerospace Engineering and The Fischell Department of Bioengineering, University of Maryland, 3188 G. L. Martin Hall, College Park, MD 20742 (aflatau@umd.edu).

Published Online: January 21, 2016. doi:10.1001/jamaophthalmol.2015.5667.

Author Contributions: Dr Flatau had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Flatau, Volpe, Quigley.

Acquisition, analysis, or interpretation of data: Flatau, Solano, Idrees, Volpe, Damion, Quigley.

Drafting of the manuscript: Flatau, Solano, Jefferys, Quigley.

Critical revision of the manuscript for important intellectual content: Flatau, Solano, Jefferys, Quigley.

Statistical analysis: Jefferys.

Administrative, technical, or material support: Flatau.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Flatau reported having a patent for a protective sleep eye shield that is intended to minimize mechanical force on the eye during sleep. Dr Quigley reported serving as a scientific consultant to SENSIMED AG, the maker of the Triggerfish device used herein. No other disclosures were reported.

Funding/Support: Contributions by Dr Flatau and Messrs Volpe and Damion are based on work supported in part by grant 13377502 from the National Science Foundation under program manager T. C. Conway, PhD. Drs Solano, Idrees, and Quigley and Ms Jefferys were supported in part by core facility grant EY01765 from the National Eye Institute (to Wilmer Eye Institute) and by unrestricted support from William T. Forrester and from Livingston and Saranne Kosberg.

Role of the Funder/Sponsor: This support allowed the coauthors to make the indicated contributions to the design and conduct of the study; supported the collection, management, analysis, and interpretation of the data; and funded the preparation and review of the manuscript for publication.

Disclaimer: Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

Additional Contributions: Jessica Shackelford, COA, and Christopher Lee, BA, COA, provided technical assistance. Both are affiliated with the Glaucoma Center of Excellence at Wilmer Eye Institute, and neither received compensation outside of their usual salary. We acknowledge the supply of some instrumentation and data from SENSIMED AG, although all data and analyses included herein were obtained and performed by the authors alone.

References
1.
Quigley  HA, Broman  AT.  The number of people with glaucoma worldwide in 2010 and 2020.  Br J Ophthalmol. 2006;90(3):262-267.PubMedGoogle ScholarCrossref
2.
Boland  MV, Quigley  HA.  Risk factors and open-angle glaucoma.  J Glaucoma. 2007;16(4):406-418.PubMedGoogle ScholarCrossref
3.
Bengtsson  B, Heijl  A.  Diurnal IOP fluctuation: not an independent risk factor for glaucomatous visual field loss in high-risk ocular hypertension.  Graefes Arch Clin Exp Ophthalmol. 2005;243(6):513-518.PubMedGoogle ScholarCrossref
4.
De Moraes  CG, Juthani  VJ, Liebmann  JM,  et al.  Risk factors for visual field progression in treated glaucoma.  Arch Ophthalmol. 2011;129(5):562-568.PubMedGoogle ScholarCrossref
5.
Nouri-Mahdavi  K, Hoffman  D, Coleman  AL,  et al; Advanced Glaucoma Intervention Study.  Predictive factors for glaucomatous visual field progression in the Advanced Glaucoma Intervention Study.  Ophthalmology. 2004;111(9):1627-1635.PubMedGoogle ScholarCrossref
6.
Heijl  A, Leske  MC, Bengtsson  B, Hyman  L, Bengtsson  B, Hussein  M; Early Manifest Glaucoma Trial Group.  Reduction of intraocular pressure and glaucoma progression.  Arch Ophthalmol. 2002;120(10):1268-1279.PubMedGoogle ScholarCrossref
7.
Quigley  HA, Addicks  EM, Green  WR, Maumenee  AE.  Optic nerve damage in human glaucoma, II.  Arch Ophthalmol. 1981;99(4):635-649.PubMedGoogle ScholarCrossref
8.
Quigley  HA, Cone  FE.  Development of diagnostic and treatment strategies for glaucoma through understanding and modification of scleral and lamina cribrosa connective tissue.  Cell Tissue Res. 2013;353(2):231-244.PubMedGoogle ScholarCrossref
9.
Norman  RE, Flanagan  JG, Sigal  IA, Rausch  SM, Tertinegg  I, Ethier  CR.  Finite element modeling of the human sclera.  Exp Eye Res. 2011;93(1):4-12.PubMedGoogle ScholarCrossref
10.
Sigal  IA, Yang  H, Roberts  MD, Burgoyne  CF, Downs  JC.  IOP-induced lamina cribrosa displacement and scleral canal expansion.  Invest Ophthalmol Vis Sci. 2011;52(3):1896-1907.PubMedGoogle ScholarCrossref
11.
Girard  MJ, Suh  JK, Bottlang  M, Burgoyne  CF, Downs  JC.  Scleral biomechanics in the aging monkey eye.  Invest Ophthalmol Vis Sci. 2009;50(11):5226-5237.PubMedGoogle ScholarCrossref
12.
Broman  AT, Quigley  HA, West  SK,  et al.  Estimating the rate of progressive visual field damage in those with open-angle glaucoma, from cross-sectional data.  Invest Ophthalmol Vis Sci. 2008;49(1):66-76.PubMedGoogle ScholarCrossref
13.
Kim  KN, Jeoung  JW, Park  KH, Lee  DS, Kim  DM.  Effect of lateral decubitus position on intraocular pressure in glaucoma patients with asymmetric visual field loss.  Ophthalmology. 2013;120(4):731-735.PubMedGoogle ScholarCrossref
14.
Korenfeld  MS, Dueker  DK.  Occult intraocular pressure elevations and optic cup asymmetry.  Invest Ophthalmol Vis Sci. 1993;34(suppl):. Abstract 1433.Google Scholar
15.
Flatau  AB.  Mitigation of the IOP elevation associated with non-supine sleep positions.  Acta Ophthalmologica.2011;89(suppl 248): Abstract 2256.Google Scholar
16.
Lee  SY, Jeoung  JW, Kim  DM, Park  KH.  Relationship between sleeping position and asymmetric visual-field loss in glaucoma patients by quantitative measurement of body position using a mobile device  [ARVO Abstract 112].  Invest Ophthalmol Vis Sci. 2015;56(7):112.Google Scholar
17.
Kim  KN, Jeoung  JW, Park  KH, Kim  DM, Ritch  R.  Relationship between preferred sleeping position and asymmetric visual field loss in open-angle glaucoma patients.  Am J Ophthalmol. 2014;157(3):739-745.PubMedGoogle ScholarCrossref
18.
Volpe  P, Flatau  A. Models of stress, strain deformation and IOP caused by non-supine sleep-position–induced loads. In: Proceedings of the BMES Frontiers in Medical Devices Conference; May 18-20, 2015; Washington, DC. Track 4, Abstract P15.
19.
Iacangelo  A, Ko  H, Kozlowski  Y, Prendergast  M, Zubajlo  R, Flatau  A. Measuring and modeling facial features for protective eye shield design. In: Proceedings of the BMES Frontiers in Medical Devices Conference; May 18-20, 2015; Washington, DC. Track 3, Abstract P73.
20.
Meurs  I, Thepass  G, Lemij  HG, Stuij  A.  Is a pillow a risk factor for glaucoma?  [ARVO Abstract 2024].  Invest Ophthalmol Vis Sci. 2015;56(7):2024.Google Scholar
21.
Mansouri  K, Medeiros  FA, Tafreshi  A, Weinreb  RN.  Continuous 24-hour monitoring of intraocular pressure patterns with a contact lens sensor.  Arch Ophthalmol. 2012;130(12):1534-1539. Medline:PubMedGoogle ScholarCrossref
22.
Lorenz  K, Korb  C, Herzog  N,  et al.  Tolerability of 24-hour intraocular pressure monitoring of a pressure-sensitive contact lens.  J Glaucoma. 2013;22(4):311-316.PubMedGoogle ScholarCrossref
23.
Mansouri  K.  The road ahead to continuous 24-hour intraocular pressure monitoring in glaucoma.  J Ophthalmic Vis Res. 2014;9(2):260-268.PubMedGoogle Scholar
24.
Mansouri  K, Shaarawy  T.  Continuous intraocular pressure monitoring with a wireless ocular telemetry sensor.  Br J Ophthalmol. 2011;95(5):627-629.PubMedGoogle ScholarCrossref
25.
De Smedt  S, Mermoud  A, Schnyder  C.  24-Hour intraocular pressure fluctuation monitoring using an ocular telemetry Sensor.  J Glaucoma. 2012;21(8):539-544.PubMedGoogle ScholarCrossref
26.
Aptel  F, Tamisier  R, Pépin  JL,  et al.  Hourly awakening vs continuous contact lens sensor measurements of 24-hour intraocular pressure.  JAMA Ophthalmol. 2014;132(10):1232-1238.PubMedGoogle ScholarCrossref
27.
Agnifili  L, Mastropasqua  R, Frezzotti  P,  et al.  Circadian intraocular pressure patterns in healthy subjects, primary open angle and normal tension glaucoma patients with a contact lens sensor.  Acta Ophthalmol. 2015;93(1):e14-e21. doi:10.1111/aos.12408.PubMedGoogle ScholarCrossref
28.
Leonardi  M, Pitchon  EM, Bertsch  A, Renaud  P, Mermoud  A.  Wireless contact lens sensor for intraocular pressure monitoring.  Acta Ophthalmol. 2009;87(4):433-437.PubMedGoogle ScholarCrossref
29.
Mansouri  K, Weinreb  RN, Liu  JH.  Efficacy of a contact lens sensor for monitoring 24-h intraocular pressure related patterns.  PLoS One. 2015;10(5):e0125530. doi:10.1371/journal.pone.0125530.PubMedGoogle ScholarCrossref
30.
World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.  JAMA. 2013;310(20):2191-2194. doi:10.1001/jama.2013.281053.PubMedGoogle ScholarCrossref
31.
Silver  DM, Geyer  O.  Pressure-volume relation for the living human eye.  Curr Eye Res. 2000;20(2):115-120.PubMedGoogle ScholarCrossref
32.
Chandrupatla  TR, Osler  TJ.  The perimeter of an ellipse.  Math Scientist. 2010;35(2):122-131.Google Scholar
33.
Wolfram Demonstrations Project. Bryant  J. Ellipsoid. http://demonstrations.wolfram.com/Ellipsoid/. Accessed July 17, 2015.
34.
Coudrillier  B, Tian  J, Alexander  S, Myers  KM, Quigley  HA, Nguyen  TD.  Biomechanics of the human posterior sclera.  Invest Ophthalmol Vis Sci. 2012;53(4):1714-1728.PubMedGoogle ScholarCrossref
35.
Hommer  A, Fuchsjäger-Mayrl  G, Resch  H, Vass  C, Garhofer  G, Schmetterer  L.  Estimation of ocular rigidity based on measurement of pulse amplitude using pneumotonometry and fundus pulse using laser interferometry in glaucoma.  Invest Ophthalmol Vis Sci. 2008;49(9):4046-4050.PubMedGoogle ScholarCrossref
36.
Downs  JC, Suh  JK, Thomas  KA, Bellezza  AJ, Hart  RT, Burgoyne  CF.  Viscoelastic material properties of the peripapillary sclera in normal and early-glaucoma monkey eyes.  Invest Ophthalmol Vis Sci. 2005;46(2):540-546.PubMedGoogle ScholarCrossref
37.
Nguyen  C, Cone  FE, Nguyen  TD,  et al.  Studies of scleral biomechanical behavior related to susceptibility for retinal ganglion cell loss in experimental mouse glaucoma.  Invest Ophthalmol Vis Sci. 2013;54(3):1767-1780.PubMedGoogle ScholarCrossref
38.
Kimball  EC, Nguyen  C, Steinhart  MR,  et al.  Experimental scleral cross-linking increases glaucoma damage in a mouse model.  Exp Eye Res. 2014;128:129-140.PubMedGoogle ScholarCrossref
39.
Gordon  MO, Beiser  JA, Brandt  JD,  et al.  The Ocular Hypertension Treatment Study.  Arch Ophthalmol. 2002;120(6):714-720.PubMedGoogle ScholarCrossref
×