A C TA 0 P H T H A L M 0 L O G I C A
69 (1991) 737-743
Fatty layer of the precorneal film in the ’office eye syndrome’ Carsten Franck National Institute of Occupational Health and Department of Occupational Medicine, Glostrup Hospital, University of Copenhagen, Denmark
Abstract. The thickness of the fatty layer of the precorneal film was estimated by the semiquantitative inteference method in 70 office workers from two town halls in Copenhagen County and compared with that in 110 controls from the general population. There were significant correlations between use of eye make-up and a thin fatty layer ( S 50 nm, P = 0.023) and between investigation time before 12 h and a thick fatty layer (2150 nm, P= 0.046). After correction for these confounders, the fatty layer was significantlyreduced in the office population compared with the controls (P=0.0032). Furthermore, absence of biomicroscopically dry eyes (expressed as a combination of premature break-up of the tear film and lissamine green-stained epithelial damage of the bulbar conjunctiva) was significantly correlated to a thick fatty layer (2 150 nm, P= 0.015), indicating that a thick fatty layer may protect the eyes against development of dry eyes. The results indicate that deficiency in available Meibomina oil is involved in the dry eye syndrome we earlier found to be associated with eye irritation in the office environment and termed the ‘office eye syndrome’. Together with observation of foam in the eye canthus, estimation of break-up time, and examination for clusters of epithelial damage on the bulbar conjunctiva, measurement of the fatty layer appears to be an easy tool for diagnosing the ‘o86ce eye syndrome’ or for differentiating this disorder from other conditions with complaints of dryness or irritation. Key words: fatty layer - semiquantitative interference study - dry eyes - tear film stability - break-up time - conjunctival epithelial damage - sick building syndrome.
In a previous report we showed that self-reported eye irritation in four town halls where indoor cli47 Acta Ophthal. 69.6
mate measurements were performed, correlated significantly with dry eyes (expressed as a combination of premature break-up of the precorneal film and clusters of epithelial damage on the bulbar conjunctiva ( P < 0.0001) (Franck 1986). The name ‘office eye syndrome’ was proposed for this new facet of the dry eye syndrome until the causes of the pathological changes have been demonstrated. In a recent study we demonstrated a decreased occurrence of foam at the inner eye canthus in the same study group compared with control persons from the general population in the same area (Franck 8c Skov 1989). Furthermore, decreased occurrence of foam was intercorrelated to both selfreported eye irritation and dry eyes. As the objective changes resemble those experimentally produced by instillation of silicone emulsion or benzalkonium chloride in the human eye, we postulated that the office environment may have a surface-active/lipophilic effect on the eyes and in this way produce a physiochemical eye dryness. Besides destroying foam (Norn 1963),reducing break-up time and causing clusters of epithelial damage (Norn & Opauszki 1977), silicone also reduces the fatty layer of the precorneal film visualized using the interference phenomenon (Norn 1979). For this reason, a follow-upstudy on eye dryness in the study population from two of the abovementioned town halls also included an investigation of the fatty layer of the precorneal film with the semiquantitativeintererence method. 737
The present study describes the correlation between the thickness of the fatty layer and dry eyes in the office eye syndrome. Further, the thickness of the fatty layer in the office population was compared with that in controls from the general population in the same area.
Subjects and Methods The study population comprised the 90 office workers from town hall I and I1 studied for eye dryness in the beginning of 1985, i.e. one year before the present investigation (Franck 1986; Franck & Skov 1989). They represented a random but stratified sample of subjects with and without eye irritation from the two town halls. Subjects with known eye diseases or use of contact lenses were excluded. Twenty persons did not participate in the present study, 9 because they had retired from active work, and 9 because of sickness or day oE,only 2 did not wish to participate in the investigation.The degree of eye dryness in participants and non-participants was not significantly different according to the results obtained early in 1985 (P= 0.62). The control group comprised 110 persons from the Glostrup Population Studies on the prevalence of cardio-vascular symptoms in an average Danish population. The participants in these studies were drawn randomly Gom the Central Person Register
among all persons living in the western part of Copenhagen County and aged 30, 40, 50 and 60 years. Our control group originally consisted of 115 subjects called in for the cardiovascularinvestigation on 10 consecutive days ultimo 1986. All were willing to join the present investigation,but 5 were excluded because of known eye disease or use of contact lenses. For the estimation of the prevalence of different values of the fatty layer in the office population the fact that the sample was weighed with regard to eye complaints was considered. The distribution of age, sex, use of eye make-up, and smoking habits is given in Table 1. Ophthalmological examinations
The investigations were carried out in the town halls. When the examination of the fatty layer was performed, the examiner did not know the results of the examination performed early in 1985or the results of the follow-up investigation on eye dryness. The follow-up investigation on eye dryness (expressed as a combination of premature breakup of the precorneal film and lissamine greenstained epithelial damage on the bulbar conjunctiva) (Franck 1986)was performed for two reasons: 1) To investigate the individual status of eye dryness after one year (to be published) and 2) to ensure that the results on the fatty layer were related to present degrees of eye dryness. Estimation of
Table 1. Characterization of study population. Office workers (N = 70)
Controls (N= 110) Significance
N
738
YO
N
Oh
the fatty layer was performed about one week after the investigationfor dryness to avoid any influence on the result of the vital stainiig used. Fatty layer
The thickness of the precorneal fatty layer was estimated on the right eye by the semiquantitative interference method (Norn 1979) using Nikon FS-2 slit-lamp. The examination was done in a semi-dark room. The subject looked fixedly just above the light spot. Diffuse light was achieved by a wide open slit and with a matt glass mounted in front of the mirror. The microscope was focussed exactly on the anterior surface of the precorneal film at magnification x 16. If red petrol-like patterns appear on the precorneal film immediately, with the subject keeping the eye open, the fatty layer is of maximal thickness and is recorded as 2 200 nm. If the fatty layer is colourless, the observer places a finger on the outside of the subject's lower lid and moves this slowly upwards in steps 1/4 of the eye opening until the red interference phenomenon is seen. The finger was placed very gently and below the tarsus to avoid expression of additional fatty on moving the lower lid. If the interference phenomenon occurs at step 1/4, the thickness is recorded as 150 nm. Likewise, occurrence at step 214 and 314 is recorded as 100 and 50 nm, respectively. If no interference phenomenon occurs, the thickness is recorded as 0 nm. A value of 0 nm is only accepted if a subsequent wide opening of the eye does not show the red pattern. This last procedure is done to avoid that an excessive fatty content should be the reason for the missing interference phenomenon (Norn 1983). In these cases a silky appearance of the fatty layer also helps to avoid this error. Objective dryness index
As an objective measure of dry eyes the subjects were divided into 3 groups according to break-up time and epithelial damage on the conjunctiva. Break-up time was measured after instillation of 10 ~ 1 1 fluorescein, % and epithelial damage as spots of jointly stained dots on the bulbar conjunctiva, nasally and temporally after instillation of 10 pl 1% lissamine green dye (Franck 1976). If there were more than one spot ofjointly stained dots, the subjects was grouped according to the largest spot. The following categorization of dry eye was 47*
4
0 50 100 150 Thickness of the fatty layer (nm)
2200
Fig 1. The frequency of the different thicknesses of the fatty layer in the total investigationpopulation divided into A: females (N = 105), B: females not using eye make-up (N = 46), C: females using eye make-up (N = 59) and D: males (N = 75). Statistical analysis using Wilcoxon rank sum test: A against D, P = 0.14. B against C, P = 0.082. C against B + D, P = 0.024.
used: No dryness: break-up time of more than 10 sec and less than 10 lissamine green-stained dots. Borderline dryness: a break-up time in the interval 5 to 10 sec and/or 10-50 lissamine green stained dots. Dryness: a break-up time below 5 sec and/or more than 50 lissamine green-stained dots. In previous studies, grouping of the two dry eye tests in an objective dryness index showed the latter index to have a highly statistical correlation to both self-reported eye irritation (Franck 1986)and occurrence of foam in the medial eye canthus (Franck & Skov 1989).
Results Table 1 shows the demographic data of the two study populations. The distribution of age, sex and smoking habits showed no statistical differences, but significantlymore females in the office population used eye make-up (P=O.O026) and significantly more office workers were investigated for thickness of the fatty layer before 12 h ( P < 0.0001). In Fig. 1 the frequency of the different thickness 739
5 50
Confounder
nm (0-50 vs 100-200 nm)
N
Statistics: Chi-square test.
N
%
(0-100 vs 150-200 nm)
O h
P
N
%
N
%
P
* Only females N = 105.
of the fatty layer is shown for the total study population divided into males, females, and females using or not using eye make-up. The majority of the subjects had a fatty layer of 50 or 100 nm. The majority of females using eye make-up had a fatty layer of 50 nm, whereas it was 100 nm in the majority of females not using eye make-up. The statistical analysis of the distributions of the different
Table 3. Frequency of a lipid layer 5 50 nm in relation to use of eye make-up and a lipid layer t 150 nm in relation to an investigation time before 12 h. Office workers (N = 70)
Controls (N=110)
Frequency of a lipid layer 5 50 nm Females 20 66.7 Using eye make-up Notusingeyemake-up 5 45.5
15 12
51.7 34.3
Frequency of a lipid layer 1150 nm Investigation time 8 16.3 Before 12 h 1 4.8 After 12 h
11 8
26.2 11.8
Statistics: Chi-square test: No significant differences.
740
t 150 nm
Category
frequencies indicate an effect of eye make-up on the thickness of the fatty layer when the females are analysed separately, and a significant effect when the females using eye make-up are tested against males plus females not using eye make-up. Table 2 shows the influence of possible confounders on a fatty layer s 50 nm or > 150 nm in the total study population. No significant effects of age, sex, or smoking habits were found, but significantly more females using eye make-up had a fatty laery I 50 nm and significantly more subjects investigated before 12 h had a fatty layer > 150 nm. Although not significant, the same trends were found in both populations (Table 3). Furthermore, a logistic regression analysis was performed on the data in Table 2. Age (544 vs > 45 years),sex, use of eye make-up (males categorized as non-users),and the investigation time were entered into the model. The results showed that only use of eye make-up correlated significantlywith a fatty layer I50 nm and that only an investigation time before 12 h correlated with a fatty layer 2 150 nm. Thus, the results were similar to those obtained with the Chi-square test. Fig. 2 shows the frequency of the different thickness of the fatty layer in the office workers compared with controls. As a whole, the fatty layer in the office workers was significantly reduced compared with that in the controls using Wilcoxon
r 40
30
x
h
v
10
"
0 50 100 150 Thickness of the fatty layer (nm)
2200
Fig. 2. The frequency of the different thicknesses of the fatty layer in A controls (N = 110), B: office workers (N = 70) and C: office workers adjusted for the effects on the distribution of the fatty layer thickness caused by the overfrequency of females using eye make-up and of subjects investigated before 12 h in the office population (N = 70). Statistical analysis using Wdcoxon rank sum test: A against B, P=0.0042 and A against C, P=0.0032.
founders are opposite in their effects on the fatty layer. We also performed a stepwise analysis of the different frequenciesof the fatty layer thickness in the two populations as a comparison of two binomial distributions, one describing a thickness of the fatty layer greater than or equal to a certain threshold level, the other describing a thickness of the fatty layer smaller than this threshold level. The result of this analysis showed that the difference in a fatty layer of 0 nm vs 50-200 nm was significant (P= 0.0012). The difference 0-50 nm vs 100-200nm was significant too (P=O.OOlO), whereas 0-100 nm vs 150-200nm or 0-150 n m vs 200 nm was not significant ( P = 0.27 and P= 0.81, respectively). Table 4 shows the correlation between dry eyes and a fatty layer < 50 nm or 2 150 nm in the office workers. A significant number of subjects without dry eyes had a fatty layer 2 150 nm (P=0.015). Because of the influence of the investigation time on the thickness of the fatty layer, we also listed the number of subjects investigated before 12 h (Table 4). Only 63% of the subjects without dryness were investigated before 12 h, whereas 75%of the subjects with dryness or borderline dryness were investigated before 12 h.
rank sum test for statistical analysis (P=O.O042),
also after correction for the effect on the fatty layer caused by the overfrequency of females using eye make-up in the office population and for the effect on the fatty layer caused by the overfrequency of office workers investigated.before12 h. The correction was performed in two steps. After correction for make-up use, using the distribtions shown in Fig. 1, the significancediminished to P= 0.011, but after correction for the investigation time, the significance was P=O.O032, because the two con-
Discussion Using the interference method, Norn originally found that the interference phenomenon occurs in most normal subjects when the eyes are half closed, corresponding to a fatty layer thickness of about 100 nm (Norn 1979)or of 67 nm after correction for the fatty layer refractive index of 1.5 (Olsen 1985). Only 5% showed a maximum thickness of the fatty layer and 8%a minimum thickness.This is
Table 4. Correlation between dry eyes and a lipid layer < 50 nm or 2 150 nm in the office workers (N = 70).
Dry eye index
Investigated before 12 h N
Yo
t 150 nm (0-100 vs 150-200 nm)
5 5 0 nm (0-50 vs 100-200 nm)
N
%
N
%
P
N
%
N
%
P
Statistics:Wdcoxon rank sum test.
741
in agreement with our results born the general population where 6% showed a fatty layer 2200 nm, 6% a minimum thickness, and the majority a layer of 100 nm. Also the missing correlation to age and sex agrees with the results obtained by Norn. The reduced fatty layer in the office population supports our earlier assumption that the ofice environment may have a surface-active/lipophilic action on the eyes, with decreased occurrence of foam and increased eye dryness, as manifested by premature break-up of the precorneal film and clusters of epithelial damage on the bulbar conjunctiva (Franck 1986; Franck & Skov 1989).The effect of silicone is an experimental model for the action of surface-active compounds on the external eye, since silicone is a very potent surface-active compound with a surface tension as low as 20 mNm-'. The low surface tension eliminatessurface elasticity, the property that is responsible for the duration of foams (Florence & Attwood 1986).Silicone displaces foam stabilisers, e.g. the phospholipids in the fatty layer. In this way silicone is able to disrupt the fatty layer into numerous oil droplets which are then suspended. After instillation of silicone in the eye, a new fatty layer does not begin to be restored until after 5-10 min (Norn 1979). The disruption of the fatty layer induces a surface tension gradient which makes the tear film unstable and causes premature break-up (Holly & Lemp 1977). The cluster of epithelial damage may be a consequence of dry spots or fatty contamination of the tear epithelium interface. Similar mechanisms are described for lipophilic surface-active compounds, like benzalkonium chloride, used as preservatives in some topical drugs (Holly 1978). Besides destroying the fatty layer, these compounds are able to induce a chronically unstable tear film, probably by interfering with the microvilli on the corneal epithelium (Holly & Lemp 1977; Pfister & Burstein 1976). Such a mechanism could explain preliminary results from our laboratory, indicating that the unstable tear film and epithelial damage in the office population to a great extent are chronic. More recently we have obtained experimental evidence that lipophilic compounds could play a role in the development of eye irritation in modem office buildings. Under controlled conditions in a climate chamber, exposure of humans to nylon carpets or rubber floor coverings, both of which emit demonstrable amounts of highly lipophilic volatile organic compounds in742
cluding undecane (Wolkoff et al. 1991).In addition to the subjectively reported eye irritation, biomicroscopical examinations showed significantly reduced foam in the medial eye canthus, reduced precorneal fatty layer, and reduced break-up time compared with exposure to an empty chamber (Johnsen et al. 1991). Use of eye make-up was found to significantly reduce the fatty layer. This is in agreement with our earlier observation that eye make-up also reduces foam formation in the medial eye canthus. Both the fatty precorneal film and foam in the eye canthus consist of fat, most likely from the Meibomian glands. Oils in the eye make-up may be able to displace the phospholipids and in this way influence both foam formation and the fatty precorneal layer. Originally Norn observed on himself that the fatty layer of the precorneal film was 2200 nm when he opened his eyes in the morning after sleep (Norn 1979).Further, he noted that the fatty layer returned to normal between 20 and 60 min later and remained normal during the rest of the day. Our results show that the fatty layer decreases in the course of the day, as among the subjects examined before 12 h the number with a fatty layer 2 150 nm was significantlyhigher than that among those examined after 12 h. In the controls, the fall in the frequency of subjects with a fatty layer 2 150 nm after 12 h was accompanied by an increased fi-equency of a fatty layer of 100 nm, but not in the frequency of a fatty layer 550 nm. In the office population, there was a fall after 12 h in the frequency of both a fatty layer t 150 nm and of 100 nm, together with an increased frequency of a fatty layer 5 50 nm. Although not significant (I"0.09), the result suggest that in the course of the day the fatty layer decreases faster in office workers than in controls. The statistically si&icant negative correlation between a fatty layer 2 150 nm and different degrees of dry eyes indicates that a thick fatty layer protects the eyes against development of dry eyes. In other words, a thick fatty layer may enable the eye to cope with surface-active/lipophilic compounds in the office environment and in that way protect the eye against development of the dry eye syndrome. Together with observation of foam in the medial eye canthus, estimation of break-up time, and examination for clusters of lissamine green-
stained epithelial damage, estimation of the fatty layer with the interference method appears to be an easy diagnostic tool in the ‘office eye syndrome’ and for differentiating this disorder from other conditions with complaints of dryness or irritation, like keratoconjunctivitis sicca or conjunctivitis simplex. Both in keratoconjunctivitis sicca and the ‘office eye syndrome’, the break-up time is markedly reduced (Norn 1969; Franck 1986),but in patients with keratoconjunctivitis sicca the precorneal fatty layer is increased and 40% of the patients have a fatty layer 2200 nm (Norn 1979). In conjunctivitis simplex the fatty layer is normal (Norn 1979),but this condition has been found associated with an increased amount of foam in the eye canthus (Norn 1963).In contrast, in the ‘officeeye syndrome’ foam is almost absent (Franck & Skov 1989).
Acknowledgment The author thanks Professor Mogens Norn for his advice and practical help in setting up the method.
References Florence A T & Attwood D (1986):Physicochemical Principles of Pharmacy, pp 271-272. Macmillan Education Ltd, London. Franck C (1986): Eye symptoms and signs in buildings with indoor climate problems (‘office eye syndrome’). Acta Ophthalmol (Copenh) 6 4 306-311. Franck C & Skov P (1989):Foam at inner eye canthus in office workers, compared with an average Danish population as control group. Acta Ophthalmol (Copenh) 67: 61-68. Holly F J (1978):Surface chemical evaluation of artificial tear and their ingredients. 11. Interfacial activity at equilibrium. Contemp Intraoc Lens Med J 4 52-65.
Holly F J & Lemp M A (1977): Tear physiology and dry eyes. Surv Ophthalmol22: 69-87. Johnsen C R, Heinig J, Schmidt K, Albrechtsen 0, Nielsen P A, Wolkoff P, Nielsen G D, Hansen L F & Franck C (1991):Controlled human reactions to building materials in climatic chambers. Part I: Performance and comfort. Indoor Air 1991 (in press). Norn M (1963): Foam at outer palpebral canthus. Acta Ophthalmol (Copenh) 41: 531-537. Norn M (1969):Desiccation of the precorneal film I. Corneal wetting time. Acta Ophthalmol (Copenh) 47: 865-880. Norn M (1979): Semiquantitastive interference study of lipid layer of precorneal film. Acta Ophthalmol (Copenh) 57 766-774. Norn M (1983): External Eye. Methods of Examination, pp 82-84, Scriptor, Copenhagen. Norn M & Opauszki A (1977): Effects of ophthalmic vehicles on the stability of the precorneal film. Acta Ophthalmol (Copenh) 55: 24-34. Olsen T (1985): Reflectometry of the precorneal film. Acta Ophthalmol (Copenh) 63: 432-438. Pfister R R & Burstein N L (1976):The effects ophthalmic drugs, vehicles, and preservatives on corneal epithelium: A scanning electron microscopic study. Invest Ophthalmol Vis Sci 15: 246-259. Wolkoff P, Nielsen G D, Hansen L F, Albrechtsen 0, Johnsen C R, Heinig J, Franck C & Nielsen P A (1991): Controlled human reactions to building materials in climatic chambers.Part Ik VOC measurements, mouse bioassay and dicipol evaluation in the 1-2 pg/m3 TVOC range. Indoor Air 1991 (in press).
Received on May 8th, 1991. Author’s address:
Carsten Franck, Juelsvej 15, DK-4390 Vippercad, Denmark.
743
69 (1991) 737-743
Fatty layer of the precorneal film in the ’office eye syndrome’ Carsten Franck National Institute of Occupational Health and Department of Occupational Medicine, Glostrup Hospital, University of Copenhagen, Denmark
Abstract. The thickness of the fatty layer of the precorneal film was estimated by the semiquantitative inteference method in 70 office workers from two town halls in Copenhagen County and compared with that in 110 controls from the general population. There were significant correlations between use of eye make-up and a thin fatty layer ( S 50 nm, P = 0.023) and between investigation time before 12 h and a thick fatty layer (2150 nm, P= 0.046). After correction for these confounders, the fatty layer was significantlyreduced in the office population compared with the controls (P=0.0032). Furthermore, absence of biomicroscopically dry eyes (expressed as a combination of premature break-up of the tear film and lissamine green-stained epithelial damage of the bulbar conjunctiva) was significantly correlated to a thick fatty layer (2 150 nm, P= 0.015), indicating that a thick fatty layer may protect the eyes against development of dry eyes. The results indicate that deficiency in available Meibomina oil is involved in the dry eye syndrome we earlier found to be associated with eye irritation in the office environment and termed the ‘office eye syndrome’. Together with observation of foam in the eye canthus, estimation of break-up time, and examination for clusters of epithelial damage on the bulbar conjunctiva, measurement of the fatty layer appears to be an easy tool for diagnosing the ‘o86ce eye syndrome’ or for differentiating this disorder from other conditions with complaints of dryness or irritation. Key words: fatty layer - semiquantitative interference study - dry eyes - tear film stability - break-up time - conjunctival epithelial damage - sick building syndrome.
In a previous report we showed that self-reported eye irritation in four town halls where indoor cli47 Acta Ophthal. 69.6
mate measurements were performed, correlated significantly with dry eyes (expressed as a combination of premature break-up of the precorneal film and clusters of epithelial damage on the bulbar conjunctiva ( P < 0.0001) (Franck 1986). The name ‘office eye syndrome’ was proposed for this new facet of the dry eye syndrome until the causes of the pathological changes have been demonstrated. In a recent study we demonstrated a decreased occurrence of foam at the inner eye canthus in the same study group compared with control persons from the general population in the same area (Franck 8c Skov 1989). Furthermore, decreased occurrence of foam was intercorrelated to both selfreported eye irritation and dry eyes. As the objective changes resemble those experimentally produced by instillation of silicone emulsion or benzalkonium chloride in the human eye, we postulated that the office environment may have a surface-active/lipophilic effect on the eyes and in this way produce a physiochemical eye dryness. Besides destroying foam (Norn 1963),reducing break-up time and causing clusters of epithelial damage (Norn & Opauszki 1977), silicone also reduces the fatty layer of the precorneal film visualized using the interference phenomenon (Norn 1979). For this reason, a follow-upstudy on eye dryness in the study population from two of the abovementioned town halls also included an investigation of the fatty layer of the precorneal film with the semiquantitativeintererence method. 737
The present study describes the correlation between the thickness of the fatty layer and dry eyes in the office eye syndrome. Further, the thickness of the fatty layer in the office population was compared with that in controls from the general population in the same area.
Subjects and Methods The study population comprised the 90 office workers from town hall I and I1 studied for eye dryness in the beginning of 1985, i.e. one year before the present investigation (Franck 1986; Franck & Skov 1989). They represented a random but stratified sample of subjects with and without eye irritation from the two town halls. Subjects with known eye diseases or use of contact lenses were excluded. Twenty persons did not participate in the present study, 9 because they had retired from active work, and 9 because of sickness or day oE,only 2 did not wish to participate in the investigation.The degree of eye dryness in participants and non-participants was not significantly different according to the results obtained early in 1985 (P= 0.62). The control group comprised 110 persons from the Glostrup Population Studies on the prevalence of cardio-vascular symptoms in an average Danish population. The participants in these studies were drawn randomly Gom the Central Person Register
among all persons living in the western part of Copenhagen County and aged 30, 40, 50 and 60 years. Our control group originally consisted of 115 subjects called in for the cardiovascularinvestigation on 10 consecutive days ultimo 1986. All were willing to join the present investigation,but 5 were excluded because of known eye disease or use of contact lenses. For the estimation of the prevalence of different values of the fatty layer in the office population the fact that the sample was weighed with regard to eye complaints was considered. The distribution of age, sex, use of eye make-up, and smoking habits is given in Table 1. Ophthalmological examinations
The investigations were carried out in the town halls. When the examination of the fatty layer was performed, the examiner did not know the results of the examination performed early in 1985or the results of the follow-up investigation on eye dryness. The follow-up investigation on eye dryness (expressed as a combination of premature breakup of the precorneal film and lissamine greenstained epithelial damage on the bulbar conjunctiva) (Franck 1986)was performed for two reasons: 1) To investigate the individual status of eye dryness after one year (to be published) and 2) to ensure that the results on the fatty layer were related to present degrees of eye dryness. Estimation of
Table 1. Characterization of study population. Office workers (N = 70)
Controls (N= 110) Significance
N
738
YO
N
Oh
the fatty layer was performed about one week after the investigationfor dryness to avoid any influence on the result of the vital stainiig used. Fatty layer
The thickness of the precorneal fatty layer was estimated on the right eye by the semiquantitative interference method (Norn 1979) using Nikon FS-2 slit-lamp. The examination was done in a semi-dark room. The subject looked fixedly just above the light spot. Diffuse light was achieved by a wide open slit and with a matt glass mounted in front of the mirror. The microscope was focussed exactly on the anterior surface of the precorneal film at magnification x 16. If red petrol-like patterns appear on the precorneal film immediately, with the subject keeping the eye open, the fatty layer is of maximal thickness and is recorded as 2 200 nm. If the fatty layer is colourless, the observer places a finger on the outside of the subject's lower lid and moves this slowly upwards in steps 1/4 of the eye opening until the red interference phenomenon is seen. The finger was placed very gently and below the tarsus to avoid expression of additional fatty on moving the lower lid. If the interference phenomenon occurs at step 1/4, the thickness is recorded as 150 nm. Likewise, occurrence at step 214 and 314 is recorded as 100 and 50 nm, respectively. If no interference phenomenon occurs, the thickness is recorded as 0 nm. A value of 0 nm is only accepted if a subsequent wide opening of the eye does not show the red pattern. This last procedure is done to avoid that an excessive fatty content should be the reason for the missing interference phenomenon (Norn 1983). In these cases a silky appearance of the fatty layer also helps to avoid this error. Objective dryness index
As an objective measure of dry eyes the subjects were divided into 3 groups according to break-up time and epithelial damage on the conjunctiva. Break-up time was measured after instillation of 10 ~ 1 1 fluorescein, % and epithelial damage as spots of jointly stained dots on the bulbar conjunctiva, nasally and temporally after instillation of 10 pl 1% lissamine green dye (Franck 1976). If there were more than one spot ofjointly stained dots, the subjects was grouped according to the largest spot. The following categorization of dry eye was 47*
4
0 50 100 150 Thickness of the fatty layer (nm)
2200
Fig 1. The frequency of the different thicknesses of the fatty layer in the total investigationpopulation divided into A: females (N = 105), B: females not using eye make-up (N = 46), C: females using eye make-up (N = 59) and D: males (N = 75). Statistical analysis using Wilcoxon rank sum test: A against D, P = 0.14. B against C, P = 0.082. C against B + D, P = 0.024.
used: No dryness: break-up time of more than 10 sec and less than 10 lissamine green-stained dots. Borderline dryness: a break-up time in the interval 5 to 10 sec and/or 10-50 lissamine green stained dots. Dryness: a break-up time below 5 sec and/or more than 50 lissamine green-stained dots. In previous studies, grouping of the two dry eye tests in an objective dryness index showed the latter index to have a highly statistical correlation to both self-reported eye irritation (Franck 1986)and occurrence of foam in the medial eye canthus (Franck & Skov 1989).
Results Table 1 shows the demographic data of the two study populations. The distribution of age, sex and smoking habits showed no statistical differences, but significantlymore females in the office population used eye make-up (P=O.O026) and significantly more office workers were investigated for thickness of the fatty layer before 12 h ( P < 0.0001). In Fig. 1 the frequency of the different thickness 739
5 50
Confounder
nm (0-50 vs 100-200 nm)
N
Statistics: Chi-square test.
N
%
(0-100 vs 150-200 nm)
O h
P
N
%
N
%
P
* Only females N = 105.
of the fatty layer is shown for the total study population divided into males, females, and females using or not using eye make-up. The majority of the subjects had a fatty layer of 50 or 100 nm. The majority of females using eye make-up had a fatty layer of 50 nm, whereas it was 100 nm in the majority of females not using eye make-up. The statistical analysis of the distributions of the different
Table 3. Frequency of a lipid layer 5 50 nm in relation to use of eye make-up and a lipid layer t 150 nm in relation to an investigation time before 12 h. Office workers (N = 70)
Controls (N=110)
Frequency of a lipid layer 5 50 nm Females 20 66.7 Using eye make-up Notusingeyemake-up 5 45.5
15 12
51.7 34.3
Frequency of a lipid layer 1150 nm Investigation time 8 16.3 Before 12 h 1 4.8 After 12 h
11 8
26.2 11.8
Statistics: Chi-square test: No significant differences.
740
t 150 nm
Category
frequencies indicate an effect of eye make-up on the thickness of the fatty layer when the females are analysed separately, and a significant effect when the females using eye make-up are tested against males plus females not using eye make-up. Table 2 shows the influence of possible confounders on a fatty layer s 50 nm or > 150 nm in the total study population. No significant effects of age, sex, or smoking habits were found, but significantly more females using eye make-up had a fatty laery I 50 nm and significantly more subjects investigated before 12 h had a fatty layer > 150 nm. Although not significant, the same trends were found in both populations (Table 3). Furthermore, a logistic regression analysis was performed on the data in Table 2. Age (544 vs > 45 years),sex, use of eye make-up (males categorized as non-users),and the investigation time were entered into the model. The results showed that only use of eye make-up correlated significantlywith a fatty layer I50 nm and that only an investigation time before 12 h correlated with a fatty layer 2 150 nm. Thus, the results were similar to those obtained with the Chi-square test. Fig. 2 shows the frequency of the different thickness of the fatty layer in the office workers compared with controls. As a whole, the fatty layer in the office workers was significantly reduced compared with that in the controls using Wilcoxon
r 40
30
x
h
v
10
"
0 50 100 150 Thickness of the fatty layer (nm)
2200
Fig. 2. The frequency of the different thicknesses of the fatty layer in A controls (N = 110), B: office workers (N = 70) and C: office workers adjusted for the effects on the distribution of the fatty layer thickness caused by the overfrequency of females using eye make-up and of subjects investigated before 12 h in the office population (N = 70). Statistical analysis using Wdcoxon rank sum test: A against B, P=0.0042 and A against C, P=0.0032.
founders are opposite in their effects on the fatty layer. We also performed a stepwise analysis of the different frequenciesof the fatty layer thickness in the two populations as a comparison of two binomial distributions, one describing a thickness of the fatty layer greater than or equal to a certain threshold level, the other describing a thickness of the fatty layer smaller than this threshold level. The result of this analysis showed that the difference in a fatty layer of 0 nm vs 50-200 nm was significant (P= 0.0012). The difference 0-50 nm vs 100-200nm was significant too (P=O.OOlO), whereas 0-100 nm vs 150-200nm or 0-150 n m vs 200 nm was not significant ( P = 0.27 and P= 0.81, respectively). Table 4 shows the correlation between dry eyes and a fatty layer < 50 nm or 2 150 nm in the office workers. A significant number of subjects without dry eyes had a fatty layer 2 150 nm (P=0.015). Because of the influence of the investigation time on the thickness of the fatty layer, we also listed the number of subjects investigated before 12 h (Table 4). Only 63% of the subjects without dryness were investigated before 12 h, whereas 75%of the subjects with dryness or borderline dryness were investigated before 12 h.
rank sum test for statistical analysis (P=O.O042),
also after correction for the effect on the fatty layer caused by the overfrequency of females using eye make-up in the office population and for the effect on the fatty layer caused by the overfrequency of office workers investigated.before12 h. The correction was performed in two steps. After correction for make-up use, using the distribtions shown in Fig. 1, the significancediminished to P= 0.011, but after correction for the investigation time, the significance was P=O.O032, because the two con-
Discussion Using the interference method, Norn originally found that the interference phenomenon occurs in most normal subjects when the eyes are half closed, corresponding to a fatty layer thickness of about 100 nm (Norn 1979)or of 67 nm after correction for the fatty layer refractive index of 1.5 (Olsen 1985). Only 5% showed a maximum thickness of the fatty layer and 8%a minimum thickness.This is
Table 4. Correlation between dry eyes and a lipid layer < 50 nm or 2 150 nm in the office workers (N = 70).
Dry eye index
Investigated before 12 h N
Yo
t 150 nm (0-100 vs 150-200 nm)
5 5 0 nm (0-50 vs 100-200 nm)
N
%
N
%
P
N
%
N
%
P
Statistics:Wdcoxon rank sum test.
741
in agreement with our results born the general population where 6% showed a fatty layer 2200 nm, 6% a minimum thickness, and the majority a layer of 100 nm. Also the missing correlation to age and sex agrees with the results obtained by Norn. The reduced fatty layer in the office population supports our earlier assumption that the ofice environment may have a surface-active/lipophilic action on the eyes, with decreased occurrence of foam and increased eye dryness, as manifested by premature break-up of the precorneal film and clusters of epithelial damage on the bulbar conjunctiva (Franck 1986; Franck & Skov 1989).The effect of silicone is an experimental model for the action of surface-active compounds on the external eye, since silicone is a very potent surface-active compound with a surface tension as low as 20 mNm-'. The low surface tension eliminatessurface elasticity, the property that is responsible for the duration of foams (Florence & Attwood 1986).Silicone displaces foam stabilisers, e.g. the phospholipids in the fatty layer. In this way silicone is able to disrupt the fatty layer into numerous oil droplets which are then suspended. After instillation of silicone in the eye, a new fatty layer does not begin to be restored until after 5-10 min (Norn 1979). The disruption of the fatty layer induces a surface tension gradient which makes the tear film unstable and causes premature break-up (Holly & Lemp 1977). The cluster of epithelial damage may be a consequence of dry spots or fatty contamination of the tear epithelium interface. Similar mechanisms are described for lipophilic surface-active compounds, like benzalkonium chloride, used as preservatives in some topical drugs (Holly 1978). Besides destroying the fatty layer, these compounds are able to induce a chronically unstable tear film, probably by interfering with the microvilli on the corneal epithelium (Holly & Lemp 1977; Pfister & Burstein 1976). Such a mechanism could explain preliminary results from our laboratory, indicating that the unstable tear film and epithelial damage in the office population to a great extent are chronic. More recently we have obtained experimental evidence that lipophilic compounds could play a role in the development of eye irritation in modem office buildings. Under controlled conditions in a climate chamber, exposure of humans to nylon carpets or rubber floor coverings, both of which emit demonstrable amounts of highly lipophilic volatile organic compounds in742
cluding undecane (Wolkoff et al. 1991).In addition to the subjectively reported eye irritation, biomicroscopical examinations showed significantly reduced foam in the medial eye canthus, reduced precorneal fatty layer, and reduced break-up time compared with exposure to an empty chamber (Johnsen et al. 1991). Use of eye make-up was found to significantly reduce the fatty layer. This is in agreement with our earlier observation that eye make-up also reduces foam formation in the medial eye canthus. Both the fatty precorneal film and foam in the eye canthus consist of fat, most likely from the Meibomian glands. Oils in the eye make-up may be able to displace the phospholipids and in this way influence both foam formation and the fatty precorneal layer. Originally Norn observed on himself that the fatty layer of the precorneal film was 2200 nm when he opened his eyes in the morning after sleep (Norn 1979).Further, he noted that the fatty layer returned to normal between 20 and 60 min later and remained normal during the rest of the day. Our results show that the fatty layer decreases in the course of the day, as among the subjects examined before 12 h the number with a fatty layer 2 150 nm was significantlyhigher than that among those examined after 12 h. In the controls, the fall in the frequency of subjects with a fatty layer 2 150 nm after 12 h was accompanied by an increased fi-equency of a fatty layer of 100 nm, but not in the frequency of a fatty layer 550 nm. In the office population, there was a fall after 12 h in the frequency of both a fatty layer t 150 nm and of 100 nm, together with an increased frequency of a fatty layer 5 50 nm. Although not significant (I"0.09), the result suggest that in the course of the day the fatty layer decreases faster in office workers than in controls. The statistically si&icant negative correlation between a fatty layer 2 150 nm and different degrees of dry eyes indicates that a thick fatty layer protects the eyes against development of dry eyes. In other words, a thick fatty layer may enable the eye to cope with surface-active/lipophilic compounds in the office environment and in that way protect the eye against development of the dry eye syndrome. Together with observation of foam in the medial eye canthus, estimation of break-up time, and examination for clusters of lissamine green-
stained epithelial damage, estimation of the fatty layer with the interference method appears to be an easy diagnostic tool in the ‘office eye syndrome’ and for differentiating this disorder from other conditions with complaints of dryness or irritation, like keratoconjunctivitis sicca or conjunctivitis simplex. Both in keratoconjunctivitis sicca and the ‘office eye syndrome’, the break-up time is markedly reduced (Norn 1969; Franck 1986),but in patients with keratoconjunctivitis sicca the precorneal fatty layer is increased and 40% of the patients have a fatty layer 2200 nm (Norn 1979). In conjunctivitis simplex the fatty layer is normal (Norn 1979),but this condition has been found associated with an increased amount of foam in the eye canthus (Norn 1963).In contrast, in the ‘officeeye syndrome’ foam is almost absent (Franck & Skov 1989).
Acknowledgment The author thanks Professor Mogens Norn for his advice and practical help in setting up the method.
References Florence A T & Attwood D (1986):Physicochemical Principles of Pharmacy, pp 271-272. Macmillan Education Ltd, London. Franck C (1986): Eye symptoms and signs in buildings with indoor climate problems (‘office eye syndrome’). Acta Ophthalmol (Copenh) 6 4 306-311. Franck C & Skov P (1989):Foam at inner eye canthus in office workers, compared with an average Danish population as control group. Acta Ophthalmol (Copenh) 67: 61-68. Holly F J (1978):Surface chemical evaluation of artificial tear and their ingredients. 11. Interfacial activity at equilibrium. Contemp Intraoc Lens Med J 4 52-65.
Holly F J & Lemp M A (1977): Tear physiology and dry eyes. Surv Ophthalmol22: 69-87. Johnsen C R, Heinig J, Schmidt K, Albrechtsen 0, Nielsen P A, Wolkoff P, Nielsen G D, Hansen L F & Franck C (1991):Controlled human reactions to building materials in climatic chambers. Part I: Performance and comfort. Indoor Air 1991 (in press). Norn M (1963): Foam at outer palpebral canthus. Acta Ophthalmol (Copenh) 41: 531-537. Norn M (1969):Desiccation of the precorneal film I. Corneal wetting time. Acta Ophthalmol (Copenh) 47: 865-880. Norn M (1979): Semiquantitastive interference study of lipid layer of precorneal film. Acta Ophthalmol (Copenh) 57 766-774. Norn M (1983): External Eye. Methods of Examination, pp 82-84, Scriptor, Copenhagen. Norn M & Opauszki A (1977): Effects of ophthalmic vehicles on the stability of the precorneal film. Acta Ophthalmol (Copenh) 55: 24-34. Olsen T (1985): Reflectometry of the precorneal film. Acta Ophthalmol (Copenh) 63: 432-438. Pfister R R & Burstein N L (1976):The effects ophthalmic drugs, vehicles, and preservatives on corneal epithelium: A scanning electron microscopic study. Invest Ophthalmol Vis Sci 15: 246-259. Wolkoff P, Nielsen G D, Hansen L F, Albrechtsen 0, Johnsen C R, Heinig J, Franck C & Nielsen P A (1991): Controlled human reactions to building materials in climatic chambers.Part Ik VOC measurements, mouse bioassay and dicipol evaluation in the 1-2 pg/m3 TVOC range. Indoor Air 1991 (in press).
Received on May 8th, 1991. Author’s address:
Carsten Franck, Juelsvej 15, DK-4390 Vippercad, Denmark.
743
