Accepted Manuscript

Impact of blinking on ocular surface and tear film parameters

Michael T.M. Wang, Leslie Tien, Alicia Han, Jung Min Lee, Dabin Kim, Maria
Markoulli, Jennifer P. Craig
PII: S1542-0124(18)30026-0
DOI: 10.1016/j.jtos.2018.06.001
Reference: JTOS 305

To appear in: Ocular Surface

Received Date: 23 January 2018

Revised Date: 3 June 2018
Accepted Date: 4 June 2018

Please cite this article as: Wang MTM, Tien L, Han A, Lee JM, Kim D, Markoulli M, Craig JP,
Impact of blinking on ocular surface and tear film parameters, Ocular Surface (2018), doi: 10.1016/
j.jtos.2018.06.001.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT
Manuscript title: Impact of blinking on ocular surface and tear film parameters

Short title: Blinking and the tear film

Authors:

Michael T. M. Wang, MBChB1

Leslie Tien, BOptom1,2

PT
Alicia Han, BOptom1,2
Jung Min Lee, BOptom1,2
Dabin Kim, BOptom1,2

RI
Maria Markoulli, PhD MOptom FAAO FBCLA3
Jennifer P. Craig, PhD MCOptom FAAO FBCLA1

SC
Corresponding author:

U
Associate Professor Jennifer P. Craig
Department of Ophthalmology
AN
New Zealand National Eye Centre
The University of Auckland, New Zealand
Private Bag 92019
M

Auckland 1142
New Zealand
D

Phone: +64 9 923 8173

Fax: +64 9 367 7173
TE

Email: jp.craig@auckland.ac.nz
EP

Author institutions:

1
Department of Ophthalmology, New Zealand National Eye Centre, The University of
Auckland, New Zealand
C

2
Department of Optometry and Vision Science, New Zealand National Eye Centre, The
AC

University of Auckland, New Zealand

3
School of Optometry and Vision Science, University of New South Wales, Sydney,
Australia

Disclosure statements: The authors have no commercial or proprietary interest in any

concept or product described in this article.

Sources of support: This research did not receive any specific grant from funding agencies
in the public, commercial, or not-for-profit sectors.
 ACCEPTED MANUSCRIPT
1 ABSTRACT

3 Purpose: To investigate the influence of blinking on tear film parameters, ocular surface

4 characteristics, and dry eye symptomology.

PT
6 Methods: A total of 154 participants were recruited in an age, gender and ethnicity-matched

7 cross-sectional study, of which 77 exhibited clinically detectable incomplete blinking, and 77

RI
8 did not. Blink rate, dry eye symptomology, tear film parameters, and ocular surface

9 characteristics were assessed in a single clinical session.

SC
10

11 Results: Overall, a higher proportion of participants exhibiting incomplete blinking fulfilled

12
U
the TFOS DEWS II dry eye diagnostic criteria (64% versus 44%, p=0.02), with an odds ratio
AN
13 (95% CI) of 2.2 (1.2-4.2) times. Participants exhibiting incomplete blinking had higher Ocular

14 Surface Disease Index scores (18±13 versus 12±9, p=0.01), and greater levels of
M

15 meibomian gland dropout (41.3±15.7% versus 27.5±14.1%, p<0.001). Furthermore, poorer

D

16 tear film lipid layer thickness, non-invasive tear film stability, expressed meibum quality,
TE

17 eyelid notching, and anterior blepharitis grades were also observed in those exhibiting

18 incomplete blinking (all p<0.05). Blink rate did not correlate significantly with any ocular
EP

19 surface parameters (all p>0.05).

20
C

21 Conclusions: Incomplete blinking was associated with a two-fold increased risk of dry eye
AC

22 disease. The greater levels of meibomian gland dropout, as well as poorer expressed

23 meibum quality and tear film lipid layer thickness, observed would suggest that incomplete

24 blinking may predispose towards the development of evaporative dry eye.

25

26 KEYWORDS

27 Blinking; dry eye; tear film; ocular surface; eyelid; meibomian gland
 ACCEPTED MANUSCRIPT
28 1. INTRODUCTION

29

30 Dry eye syndrome is a complex disorder, whereby multiple components of the lacrimal

31 functional unit can be susceptible to disruption by a wide variety of factors.[1] The blinking

32 mechanism plays an integral role in ocular surface and tear film homeostasis.[2, 3] Blinking

PT
33 drives the lacrimal pump and facilitates spreading, mixing, and distribution of tear film

34 constituents over the ocular surface.[2-4] The muscular action of the eyelids during blinking

RI
35 also stimulates the secretion of meibomian lipids that lubricate the ocular surface and inhibit

36 tear evaporation.[2-6]

SC
37

38 Previous studies have reported that the suppressed blink rate and greater tendency for

39
U
incomplete blinking during visual concentration are associated with reduced tear film stability
AN
40 and symptoms of dry eye.[7-11] Higher levels of meibomian gland dropout and shortened

41 tear film breakup times have also been observed in diseases that adversely affect blinking
M

42 patterns, including cranial nerve VII palsy and Grave’s orbitopathy.[12-15] In contrast, the
D

43 thickness of the tear film lipid layer has been reported to increase with repeated forceful
TE

44 blinking.[5] Furthermore, improvements in dry eye symptoms have been observed with the

45 use of an animation that encouraged increased blinking among computer users.[16]

EP

46

47 As a potentially modifiable behaviour, blinking training is commonly recommended as part of

C

48 the management of dry eye.[3, 4, 17] However, the relationship between clinical
AC

49 measurements of blinking patterns and dry eye severity remains to be established. This

50 age, gender, and ethnicity-matched cross-sectional study sought to explore the influence of

51 blinking patterns on tear film parameters, ocular surface characteristics and dry eye

52 symptomology.
 ACCEPTED MANUSCRIPT
53 2. MATERIALS AND METHODS

54

55 2.1. Subjects

56

57 This age, gender, and ethnicity-matched cross-sectional, observational study was conducted

PT
58 at a single-centre University of Auckland clinic between May 23, 2016 and August 25, 2017,

59 and followed the tenets of the Declaration of Helsinki and was approved by the institutional

RI
60 human participants ethics committee. Participants were required to be 18 years of age or

61 older, with no major ocular or systemic diseases (other than dry eye), no previous ocular

SC
62 surgery, no contact lens wear within 48 hours of study participation, and no use of topical or

63 systemic medications known to affect the eye. Eligible participants were enrolled after

64 providing written consent.

U
AN
65

66 A total of 154 participants were recruited into two age, gender and ethnicity-matched groups:
M

67 77 participants exhibiting clinically detectable incomplete blinking, and 77 participants who

D

68 did not. This satisfied sample size requirements calculated using PASS 2002. Non-
TE

69 parametric adjusted power calculations were made with non-invasive tear film breakup time

70 as the designated outcome, and showed that a minimum of 41 participants per group was
EP

71 required to detect a clinically significant breakup time difference of 5s, with 80% power

72 (β=0.2), and a two-sided statistical significance level of 5% (α=0.05). The SD of normal

C

73 values was estimated to be 8s.[18]

AC

74

75 2.2. Blinking Assessment

76

77 Blinking assessment was conducted by an observer independent to the investigator

78 performing ocular measurements. To prevent interference with the natural blinking pattern,

79 participants were not informed about the blinking assessment during the initial consenting

80 process, and made aware only after the completion of all clinical measurements.
 ACCEPTED MANUSCRIPT
81

82 During enrolment, the blink rate of participants was recorded manually over a one-minute

83 interval, while the participant was unaware and not engaging in any visual concentration

84 tasks or up-gaze. Incomplete blinking was deemed to be present if noted during slow motion

85 playback of a two-minute infra-red video recording collected by the Keratograph 5M (Oculus,

PT
86 Germany), or by clinical observation during history-taking conducted by the independent

87 observer (typically over a 10-minute period). Slow playback of the infra-red video recordings

RI
88 was the more sensitive method for assessing the presence of incomplete blinking, with 15 of

89 77 (19%) participants exhibiting incomplete blinking detected on infra-red video recording but

SC
90 not clinical observation, and 0 of 77 (0%) participants detected on clinical observation but not

91 infra-red video recording.

92
U
AN
93 2.3. Ocular Surface Evaluation

94
M

95 Dry eye symptoms, tear film parameters and ocular surface characteristics were assessed,
D

96 in a single clinical session, for the right eye of each participant. The clinical assessments
TE

97 were conducted in accordance to the recommendations made by the TFOS DEWS II

98 Diagnostic Methodology subcommittee, and the diagnostic test battery was used to identify
EP

99 participants with dry eye disease.[19]

100
C

101 The Ocular Surface Disease Index (OSDI) questionnaire was administered to grade the level
AC

102 of dry eye symptomology. The clinical measurements were then conducted in ascending

103 order of invasiveness in order to minimize the impact on tear film physiology for subsequent

104 tests: tear meniscus height, non-invasive tear film breakup time, tear film lipid layer grade,

105 tear film osmolarity, slit lamp examination, ocular surface staining, meibomian gland

106 expression, and infrared meibography. All participants were assessed in the same location,

107 with a mean ± SD room temperature of 21.3 ± 0.5°C and a mean ± SD relative humidity of

108 51 ± 6%.
 ACCEPTED MANUSCRIPT
109

110 The tear meniscus height, non-invasive tear film breakup time, and tear film lipid layer grade

111 were assessed using the Oculus Keratograph 5M.[19] The lower tear meniscus height was

112 determined from an infra-red tear meniscus image using the inbuilt digital callipers. Three

113 tear meniscus height measurements near the geometric centre of the lower meniscus were

PT
114 averaged. Automated non-invasive breakup time was recorded as the time taken post-blink

115 for the first distortion in the grid reflection to be detected, while the subject viewed the

RI
116 fixation target within the instrument, and was requested to refrain from blinking. Three

117 breakup time measurements were averaged in each case.[19] Tear film lipid layer grading

SC
118 was based on the modified Guillon-Keeler grading system: grade 1, open meshwork; grade

119 2, closed meshwork; grade 3, wave or flow; grade 4, amorphous; grade 5, coloured fringes;

120
U
grade 0, non-continuous layer (non-visible or abnormal coloured fringes).[20, 21]
AN
121

122 Tear film osmolarity was evaluated with a clinical osmometer (TearLab, USA), from 50nL
M

123 tear samples collected from the lower lid tear meniscus of both eyes. Two measurements
D

124 were taken, and the higher reading and interocular difference recorded.[19]
TE

125

126 Signs of anterior blepharitis and eyelid margin abnormalities, including meibomian gland
EP

127 capping, eyelid notching, and eyelid telangiectasia, were assessed by slit lamp examination.

128 Grading of the clinical features was based on the Efron grading system for contact lens
C

129 complications: grade 0, absent; grade 1, trace; grade 2, mild; grade 3, moderate; and grade
AC

130 4 severe.[22]

131

132 Lid parallel conjunctival folds were noted,[23] and sodium fluorescein and lissamine green

133 dyes then applied, in turn, to the bulbar conjunctiva in order to evaluate the localised corneal

134 and conjunctival areas of epithelial desiccation. Staining was recorded with the Oculus

135 Keratograph 5M to allow investigator masking and evaluated according to the modified

136 Oxford grading scheme,[24] where the nasal and temporal conjunctiva were each divided
 ACCEPTED MANUSCRIPT
137 into three areas and the cornea into five areas. Staining was graded from 0 to 5 according to

138 the level of confluence in each area and summed to provide a maximum score of 55. Lid

139 wiper epitheliopathy was assessed relative to Korb grading using sodium fluorescein and

140 lissamine green dyes.[25]

141

PT
142 Inferior eyelid meibomian glands underwent diagnostic expression with the Korb Meibomian

143 Gland EvaluatorTM (TearScience®, North Carolina, USA), and the meibum quality was

RI
144 graded as: grade 0, clear fluid; grade 1, slightly turbid; grade 2, thick opaque; grade 3,

145 toothpaste like; or grade 4, complete orifice blockage.[26] Infrared meibography was then

SC
146 performed using the Oculus Keratograph 5M.[19, 27] From the captured image of the

147 everted upper eyelid, visible meibomian glands were outlined and the relative area of

148
U
meibomian gland dropout calculated using Image J software. Percentage dropout was
AN
149 calculated by dividing the area with no visible glands by the entire conjunctival area.[28, 29]

150
M

151 2.4. Statistics

D

152
TE

153 Statistical analysis was performed using Graph Pad Prism version 6.02. Between-group

154 comparisons of continuous variables were performed using the paired t-test, where normal
EP

155 distribution had been confirmed by the Kolmogorov-Smirnov test (p>0.05). Non-normally

156 distributed continuous measures and ordinal data were analysed using the Wilcoxon signed-
C

157 rank test, and categorical data using Fisher’s exact test. Statistical correlation between
AC

158 ocular surface parameters and blink rate was tested using Pearson’s product-moment

159 correlation coefficient for normally distributed continuous measurements, or Spearman’s

160 rank correlation coefficient for non-normally continuous or ordinal measurements. All tests

161 were two-tailed and p<0.05 was considered significant.

162
 ACCEPTED MANUSCRIPT
163 3. RESULTS

164

165 The mean ± SD age of the 154 participants (88 females, 66 males) was 39 ± 14 years.

166 Demographic, ocular surface, and tear film characteristics of participants exhibiting and not

167 exhibiting incomplete blinking are presented in Table 1. Overall, a significantly greater

PT
168 proportion of participants exhibiting incomplete blinking fulfilled the TFOS DEWS II dry eye

169 diagnostic criteria (64% versus 44%, p=0.02, Figure 1), with an odds ratio (95% CI) of 2.2

RI
170 (1.2-4.2) times.

171

SC
172 OSDI scores were higher in participants who exhibited incomplete blinking than those that

173 did not (18±13 versus 12±9, p=0.01, Figure 2). Participants exhibiting incomplete blinking

174
U
had poorer lipid layer grades (median grade 2 versus 3, p=0.03) and non-invasive tear film
AN
175 breakup time (median 6.1s versus 9.7s, p=0.04, Figure 3). There were no significant

176 differences between groups in tear meniscus height or tear film osmolarity measurements
M

177 (all p>0.05).

D

178
TE

179 Higher levels of meibomian gland dropout were observed in participants who exhibited

180 incomplete blinking (41.3±15.7% versus 27.5±14.1%, p<0.001, Figures 4 and 5), as well as
EP

181 poorer quality of expressed meibum, and greater severity of eyelid notching and anterior

182 blepharitis (all p<0.05). Ocular surface staining, meibomian gland capping, eyelid
C

183 telangiectasia, lid parallel conjunctival folds, and lid wiper epitheliopathy grades did not differ
AC

184 significantly between groups (all p>0.05).

185

186 No significant correlations were observed between blink rate and any of the ocular surface

187 parameters (all p>0.05, Table 2).

188
 ACCEPTED MANUSCRIPT
189 4. DISCUSSION

190

191 The results of this study show that a significantly greater proportion of participants exhibiting

192 incomplete blinking fulfilled the TFOS DEWS II dry eye diagnostic criteria, with a two-fold

193 increased risk. This was associated with higher levels of meibomian gland dropout and

PT
194 poorer quality of expressed meibum. These trends are in agreement with previous studies

195 that have reported increased meibomian gland dropout among patients with disorders that

RI
196 affect the completeness of eyelid closure during blinking, including cranial nerve VII palsy

197 and Grave’s orbitopathy.[12-15] This would appear to lend support to the hypothesis that

SC
198 incomplete blinking may predispose towards the development of meibomian gland

199 dysfunction.[2-4, 21] The blinking mechanism is understood to encourage the movement of

200
U
meibomian gland secretions into the tear film.[2-6] The contraction of the muscle of Riolan
AN
201 during each blink may potentially be involved, by exerting pressure on the meibomian glands

202 and thereby facilitating the expression of meibum through the gland orifices onto the lid
M

203 margin.[30-32] It is thought that incomplete blinking may be associated with diminished lipid
D

204 flow in the meibomian glands and ducts,[2-4, 33] due to reduced involvement of the muscle
TE

205 of Riolan.[30-32] The resulting meibum stasis may potentially trigger intraductal inflammation

206 and the development of obstructive meibomian gland dysfunction.[2, 32] This, in turn, may
EP

207 lead to meibomian gland dropout over time, due to the raised pressures within the gland as a

208 result of meibum stasis.[32] Nevertheless, the cross sectional design of the current study
C

209 precludes the inference of causal relationships, and the potential for patients with meibomian
AC

210 gland dysfunction to develop changes in the eyelid margin, which might subtly interfere with

211 the blinking mechanism, cannot be excluded.

212

213 Poorer tear film stability and lipid layer quality were also observed in participants who

214 exhibited incomplete blinking. This may potentially be related to the reduced delivery of

215 meibomian lipids to the tear film with incomplete blinking.[2-4, 33] Furthermore, blinking

216 facilitates the spreading and distribution of the lipid layer across the ocular surface, and
 ACCEPTED MANUSCRIPT
217 inefficient blinking can result in poor maintenance of the lipid layer integrity.[3, 4] The higher

218 degree of meibomian gland dropout and poorer grades of meibum quality observed in this

219 participant group are also likely to contribute. Together with the increased effective interblink

220 interval associated with incomplete blinking, the impaired lipid layer quality may encourage

221 increased evaporative losses from the tear film and desiccation of the exposed ocular

PT
222 surface.[3, 4, 34] This might potentially trigger a vicious cycle of ocular surface inflammation

223 which can further exacerbate the signs and symptoms of dry eye.[2] In the current study,

RI
224 clinical signs including eyelid notching and anterior blepharitis were noted to be more

225 pronounced in participants who exhibited incomplete blinking. These clinical signs may also

SC
226 be related to the reduced clearance of debris associated with incomplete blinking.[4]

227 Furthermore, a greater severity of dry eye symptomology among this participant group was

228
U
reflected by higher OSDI scores, with the inter-group difference (6 points) exceeding
AN
229 previously established thresholds for minimal clinically important difference in mild dry eye

230 disease (≥4.5 points). [19, 35] The mean OSDI score (of 18 points) in participants exhibiting
M

231 incomplete blinking also met the TFOS DEWS II diagnostic cut-off for dry eye disease (≥13
D

232 points), but remained within normal range for participants who did not.[19]
TE

233

234 Despite inferior ocular surface staining classically being regarded as a clinical sign of
EP

235 incomplete eyelid closure,[10] no significant intergroup differences were detected in the

236 current study. In addition, tear osmolarity did not differ significantly between participant
C

237 groups. It is conceivable for the lack of intergroup differences to be related to the milder
AC

238 overall degree of dry eye severity in both participant groups, as demonstrated by the minimal

239 ocular surface staining scores and marginal mean tear osmolarity values (308 and 307

240 mOsmol/L), which were close to, but not exceeding, the TFOS DEWS II diagnostic cut-off for

241 dry eye disease (≥308 mOsmol/L).[19] Indeed, it has been increasingly acknowledged that

242 ocular surface staining lacks discriminatory power in mild-to-moderate disease, and that

243 staining may be a clinical sign indicative of more severe disease. [19, 36]

244
 ACCEPTED MANUSCRIPT
245 The potential association between incomplete blinking and the development of meibomian

246 gland dysfunction and evaporative dry eye would support current recommendations of

247 offering blinking training as part of the management of dry eye symptoms.[3, 17] Blinking

248 training may aid in the preservation of meibomian gland structure and function, as well as

249 improve lipid secretion and tear film distribution.[3, 4] Improvements in tear film lipid layer

PT
250 thickness with deliberate and repeated forceful blinks have previously been reported.[5]

251 Techniques which encourage increased blinking, such as animations for computer users,

RI
252 have also been shown to reduce dry eye symptomology.[16] Further research is required to

253 determine whether these techniques are also effective at reducing the clinical signs of dry

SC
254 eye, and preserving the structure and function of the meibomian glands.

255

256
U
Interestingly, no significant correlations were observed between the frequency of blinking
AN
257 and any of the ocular surface parameters. This contrasts with earlier studies which report

258 poorer tear film quality with decreased blink rates during visual concentration.[7, 11] The
M

259 measurement of blink rate while participants were not engaged in activities requiring visual
D

260 concentration in the current study may provide some explanation towards the discrepancies
TE

261 observed. Furthermore, in both participant groups the average blink rate greater than 20

262 minute-1, which would be considered on the higher end of the normal range.[37, 38] The
EP

263 measurement of blink rate while the participants were engaged in conversation during the

264 enrolment process may have contributed.[37] However, an alternative explanation would
C

265 suggest that the increased blink rates may potentially reflect involuntary compensatory reflex
AC

266 mechanisms associated with dry eye, which serve to restore the tear film and lubricate the

267 ocular surface.[39] Such a compensatory reflex may also contribute towards the lack of

268 association between blink rate and ocular parameters. Furthermore, whether longitudinal

269 intrinsic variability in blink rate may also contribute, in this regard, is unknown. Nevertheless,

270 the findings of the current study would suggest that assessing blink completeness may be a

271 more useful indicator of dry eye status than the measurement of blink rate, and addressing
 ACCEPTED MANUSCRIPT
272 incomplete blinking may be an important aspect of the non-pharmacological management of

273 dry eye disease and meibomian gland dysfunction.

274

275 5. Conclusions

276

PT
277 Of note, the findings of this cross-sectional study showed that incomplete blinking was

278 associated with a two-fold increased risk of dry eye disease. Furthermore, poorer levels of

RI
279 meibomian gland dropout, expressed meibum quality, and tear film lipid layer thickness were

280 observed among participants exhibiting incomplete blinking. These trends would suggest

SC
281 that incomplete blinking may predispose towards the development of evaporative dry eye.

282

283
U
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
284 6. ACKNOWLEDGEMENTS

285

286 None.

287

288

PT
289 7. FUNDING

290

RI
291 This research did not receive any specific grant from funding agencies in the public,

292 commercial, or not-for-profit sectors.

SC
293

U
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
294 REFERENCES

295

296 [1] Craig JP, Nichols KK, Akpek EK, Caffery B, Dua HS, Joo CK, et al. TFOS DEWS II

297 definition and classification report. Ocul Surf. 2017;15:276-83.

298 [2] Bron AJ, de Paiva CS, Chauhan SK, Bonini S, Gabison EE, Jain S, et al. TFOS DEWS II

PT
299 pathophysiology report. Ocul Surf. 2017;15:438-510.

300 [3] McMonnies CW. Blink efficiency: a neglected area of ocular surface disease

RI
301 management? Invest Ophthalmol Vis Sci. 2011;52:4484.

302 [4] McMonnies CW. Incomplete blinking: exposure keratopathy, lid wiper epitheliopathy, dry

SC
303 eye, refractive surgery, and dry contact lenses. Cont Lens Anterior Eye. 2007;30:37-51.

304 [5] Korb DR, Baron DF, Herman JP, Finnemore VM, Exford JM, Hermosa JL, et al. Tear film

305
U
lipid layer thickness as a function of blinking. Cornea. 1994;13:354-9.
AN
306 [6] Foulks GN, Bron AJ. Meibomian gland dysfunction: a clinical scheme for description,

307 diagnosis, classification, and grading. Ocul Surf. 2003;1:107-26.

M

308 [7] Cardona G, Garcia C, Seres C, Vilaseca M, Gispets J. Blink rate, blink amplitude, and
D

309 tear film integrity during dynamic visual display terminal tasks. Curr Eye Res.
TE

310 2011;36:190-7.

311 [8] Hirota M, Uozato H, Kawamorita T, Shibata Y, Yamamoto S. Effect of incomplete blinking
EP

312 on tear film stability. Optom Vis Sci. 2013;90:650-7.

313 [9] Portello JK, Rosenfield M, Chu CA. Blink rate, incomplete blinks and computer vision
C

314 syndrome. Optom Vis Sci. 2013;90:482-7.

AC

315 [10] Himebaugh NL, Begley CG, Bradley A, Wilkinson JA. Blinking and tear break-up during

316 four visual tasks. Optom Vis Sci. 2009;86:E106-14.

317 [11] Patel S, Henderson R, Bradley L, Galloway B, Hunter L. Effect of visual display unit use

318 on blink rate and tear stability. Optom Vis Sci. 1991;68:888-92.

319 [12] Shah CT, Blount AL, Nguyen EV, Hassan AS. Cranial nerve seven palsy and its

320 influence on meibomian gland function. Ophthal Plast Reconstr Surg. 2012;28:166-8.
 ACCEPTED MANUSCRIPT
321 [13] Takahashi Y, Kakizaki H. Meibomian Gland Dysfunction in Cranial Nerve VII Palsy.

322 Ophthal Plast Reconstr Surg. 2015;31:179-81.

323 [14] Wan T, Jin X, Lin L, Xu Y, Zhao Y. Incomplete Blinking May Attribute to the

324 Development of Meibomian Gland Dysfunction. Curr Eye Res. 2016;41:179-85.

325 [15] Kim YS, Kwak AY, Lee SY, Yoon JS, Jang SY. Meibomian gland dysfunction in Graves'

PT
326 orbitopathy. Can J Opthalmol. 2015;50:278-82.

327 [16] Nosch DS, Foppa C, Toth M, Joos RE. Blink Animation Software to Improve Blinking

RI
328 and Dry Eye Symptoms. Optom Vis Sci. 2015;92:e310-5.

329 [17] Jones L, Downie LE, Korb D, Benitez-Del-Castillo JM, Dana R, Deng SX, et al. TFOS

SC
330 DEWS II Management and Therapy Report. Ocul Surf. 2017;15:575-628.

331 [18] Wang MT, Jaitley Z, Lord SM, Craig JP. Comparison of Self-applied Heat Therapy for

332
U
Meibomian Gland Dysfunction. Optom Vis Sci. 2015;92:e321-6.
AN
333 [19] Wolffsohn JS, Arita R, Chalmers R, Djalilian A, Dogru M, Dumbleton K, et al. TFOS

334 DEWS II Diagnostic Methodology report. Ocul Surf. 2017;15:539-74.

M

335 [20] Guillon JP. Use of the Tearscope Plus and attachments in the routine examination of the
D

336 marginal dry eye contact lens patient. Adv Exp Med Bio. 1998;438:859-67.
TE

337 [21] Craig JP, Wang MT, Kim D, Lee JM. Exploring the Predisposition of the Asian Eye to

338 Development of Dry Eye. Ocul Surf. 2016;14:385-92.

EP

339 [22] Efron N, Morgan PB, Katsara SS. Validation of grading scales for contact lens

340 complications. Ophthalmic Physiol Opt. 2001;21:17-29.

C

341 [23] Nemeth J, Fodor E, Lang Z, Kosina-Hagyo K, Berta A, Komar T, et al. Lid-parallel
AC

342 conjunctival folds (LIPCOF) and dry eye: a multicentre study. Br J Opthalmol.

343 2012;96:1380-5.

344 [24] Bron AJ, Evans VE, Smith JA. Grading of corneal and conjunctival staining in the

345 context of other dry eye tests. Cornea. 2003;22:640-50.

346 [25] Korb DR, Herman JP, Greiner JV, Scaffidi RC, Finnemore VM, Exford JM, et al. Lid

347 wiper epitheliopathy and dry eye symptoms. Eye Cont Lens. 2005;31:2-8.
 ACCEPTED MANUSCRIPT
348 [26] Bron AJ, Benjamin L, Snibson GR. Meibomian gland disease. Classification and grading

349 of lid changes. Eye. 1991;5 ( Pt 4):395-411.

350 [27] Arita R, Itoh K, Inoue K, Amano S. Noncontact infrared meibography to document age-

351 related changes of the meibomian glands in a normal population. Ophthalmol.

352 2008;115:911-5.

PT
353 [28] Pult H, Riede-Pult B. Comparison of subjective grading and objective assessment in

354 meibography. Cont Lens Anterior Eye. 2013;36:22-7.

RI
355 [29] Srinivasan S, Menzies K, Sorbara L, Jones L. Infrared imaging of meibomian gland

356 structure using a novel keratograph. Optom Vis Sci. 2012;89:788-94.

SC
357 [30] Lipham WJ, Tawfik HA, Dutton JJ. A histologic analysis and three-dimensional

358 reconstruction of the muscle of Riolan. Ophthal Plast Reconstr Surg. 2002;18:93-8.

359
U
[31] Kakizaki H, Nakano T, Ikeda H, Selva D, Leibovitch I. Tarsal elastic fiber distribution: an
AN
360 anatomic study. Ophthal Plast Reconstr Surg. 2011;27:128-9.

361 [32] Knop E, Knop N, Millar T, Obata H, Sullivan DA. The international workshop on
M

362 meibomian gland dysfunction: report of the subcommittee on anatomy, physiology, and
D

363 pathophysiology of the meibomian gland. Invest Ophthalmol Vis Sci. 2011;52:1938-78.
TE

364 [33] Kawashima M, Tsubota K. Tear lipid layer deficiency associated with incomplete

365 blinking: a case report. BMC Ophthalmol. 2013;13:34.

EP

366 [34] Craig JP, Tomlinson A. Importance of the lipid layer in human tear film stability and

367 evaporation. Optom Vis Sci. 1997;74:8-13.

C

368 [35] Miller KL, Walt JG, Mink DR, Satram-Hoang S, Wilson SE, Perry HD, et al. Minimal
AC

369 clinically important difference for the ocular surface disease index. Arch Ophthalmol.

370 2010;128:94-101.

371 [36] Sullivan BD, Whitmer D, Nichols KK, Tomlinson A, Foulks GN, Geerling G, et al. An

372 objective approach to dry eye disease severity. Invest Ophthalmol Vis Sci.

373 2010;51:6125-30.

374 [37] Bentivoglio AR, Bressman SB, Cassetta E, Carretta D, Tonali P, Albanese A. Analysis

375 of blink rate patterns in normal subjects. Mov Disord. 1997;12:1028-34.

 ACCEPTED MANUSCRIPT
376 [38] Doughty MJ, Naase T. Further analysis of the human spontaneous eye blink rate by a

377 cluster analysis-based approach to categorize individuals with 'normal' versus 'frequent'

378 eye blink activity. Eye Cont Lens. 2006;32:294-9.

379 [39] Tsubota K. Tear dynamics and dry eye. Prog Retin Eye Res. 1998;17:565-96.

380

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
381 TABLES

382

383 Table 1: Demographic, tear film and ocular surface characteristics of participants exhibiting

384 and not exhibiting incomplete blinking. Data are presented as mean ± SD, median (IQR), or

385 number of participants (% of participants). Asterisks denote statistically significant values

PT
386 (p<0.05).

387

RI
Incomplete blinking
Present Absent
Demographic or ocular characteristics (n=77) (n=77) p

SC
Age (years) 39±14 39±14 -
Female gender 44 (57%) 44 (57%) -
European ethnicity 34 (44%) 34 (44%) -

U
East Asian ethnicity 38 (49%) 38 (49%) -
South Asian ethnicity 5 (6%) 5 (6%) -
AN
Blink rate (minute-1) 22±11 21±10 0.51
TFOS DEWS II dry eye diagnosis 49 (64%) 34 (44%) 0.02*
OSDI score 18±13 12±9 0.01*
M

Tear meniscus height (mm) 0.26±0.08 0.27±0.08 0.68

Tear film lipid layer grade 2 (2-4) 3 (2-4) 0.03*
Non-invasive tear film breakup time (s) 6.1 (4.5-10.7) 9.7 (5.8-15.3) 0.04*
D

Tear film osmolarity (mOsmol/L) 308±13 307±14 0.86

Interocular difference in osmolarity 7±5 6±5 0.92
TE

(mOsmol/L)
Meibomian gland dropout (%) 41.3±15.7 27.5±14.1 <0.001*
Meibomian gland capping grade 1 (1-2) 1 (0-2) 0.39
EP

Expressed meibum grade 1 (1-3) 1 (0-2) 0.04*

Eyelid notching grade 1 (0-2) 0 (0-2) 0.03*
Eyelid telangiectasia grade 0 (0-1) 0 (0-1) 0.79
C

Anterior blepharitis grade 0 (0-1) 0 (0-0) 0.02*

Sodium fluorescein staining score (out of 55) 4 (0-9) 4 (0-8) 0.51
AC

Lissamine green staining score (out of 55) 2 (0-5) 1 (0-5) 0.73

Lid parallel conjunctival folds (LIPCOF) grade 1 (0-2) 1 (0-2) 0.94
Lid wiper epitheliopathy grade 2 (0-3) 1 (0-3) 0.07
388
389
 ACCEPTED MANUSCRIPT
390 Table 2: Correlation analysis of demographic, tear film and ocular surface characteristics

391 with blink rate. Data are presented as Pearson’s product-moment correlation coefficient for

392 continuous measurements, or Spearman’s rank correlation coefficient for ordinal

393 measurements.

394
Correlation with blink rate

PT
Demographic or ocular characteristics Coefficient p
Age (years) 0.089 0.27
OSDI score -0.068 0.40

RI
Tear meniscus height (mm) 0.015 0.86
Tear film lipid layer grade -0.004 0.96

SC
Non-invasive tear film breakup time (s) -0.072 0.38
Tear film osmolarity (mOsmol/L) 0.031 0.70
Interocular difference in osmolarity (mOsmol/L) 0.062 0.45
Meibomian gland dropout (%) 0.004 0.96

U
Meibomian gland capping grade 0.016 0.85
AN
Meibum grade 0.073 0.37
Eyelid notching grade 0.033 0.69
Eyelid telangiectasia grade 0.117 0.15
Anterior blepharitis grade 0.049 0.55
M

Sodium fluorescein staining score (out of 55) 0.095 0.24

Lissamine green staining score (out of 55) 0.010 0.90
D

Lid parallel conjunctival folds (LIPCOF) grade 0.078 0.34

Lid wiper epitheliopathy grade 0.031 0.70
TE

395
396
C EP
AC
 ACCEPTED MANUSCRIPT
397 FIGURE LEGENDS

398

399 Figure 1: Proportion of participants exhibiting and not exhibiting incomplete blinking fulfilling

400 the TFOS DEWS II dry eye diagnostic criteria. Each bar represents the percentage of

401 participants fulfilling the diagnostic criteria. Error bars represent the 95% confidence interval.

PT
402 Asterisks denote statistically significant differences (p<0.05).

403

RI
404 Figure 2: Ocular surface disease index (OSDI) scores of participants exhibiting and not

405 exhibiting incomplete blinking. Each point represents the OSDI score of an individual

SC
406 participant. Bars represent the mean OSDI score. Error bars represent the standard

407 deviation. Asterisks denote statistically significant differences (p<0.05).

408
U
AN
409 Figure 3: Non-invasive tear film breakup time of participants exhibiting and not exhibiting

410 incomplete blinking. Each point represent the tear film breakup time of an individual
M

411 participant. Bars represent the median tear film breakup time. Error bars represent the
D

412 interquartile range. Asterisks denote statistically significant differences (p<0.05).

TE

413

414 Figure 4: (A) Infrared meibography of a participant not exhibiting incomplete blinking,
EP

415 showing 17% meibomian gland dropout. (B) Infrared meibography of a participant exhibiting

416 incomplete blinking, showing 44% meibomian gland dropout.

C

417
AC

418 Figure 5: Percentage meibomian gland dropout in participants exhibiting and not exhibiting

419 incomplete blinking. Each bar represents the mean meibomian gland dropout percentage.

420 Error bars represent the standard deviation. Asterisks denote statistically significant

421 differences (p<0.05).

422
 Participants fulfilling
TFOS DEWS II dry eye diagnosis (%)

0
20
40
60
80
100

AC
C EP

Present
TE
D
M
AN
ACCEPTED MANUSCRIPT

U
p = 0.02 *

SC
RI
PT

Incomplete blinking
Absent
 ACCEPTED MANUSCRIPT

p=0.01*

100

PT
RI
80

U SC
OSDI score

60

AN
M
D
40
TE
EP

20
C
AC

0
Present Absent
Incomplete blinking
 ACCEPTED MANUSCRIPT

p=0.04*

60

PT
Non-invasive breakup time (s)

50

RI
SC
40

U
AN
30 M
D
TE

20
EP
C

10
AC

0
Present Absent
Incomplete blinking
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT
p < 0.001 *

60
Meibomian gland dropout (%)

PT
RI
SC
40

U
AN
M
D
TE
20
EP
C
AC

0
Present Absent

Incomplete blinking