A C T A O P H T H A L M O L O G I C A V O L . 5 6 1 9 7 8

The Institute of Clinical Genetics, University of Odense (Head: M. Hauge), and

The Eye Department, Kommunehospitalet (the Municipal Hospital), Copenhagen (Heads: P. Brrendstrup, S. E. Lorentzen, M. S. Norn and K. Nsrskov)

CORNEAL THICKNESS

I I . Environmental and Genetic Factors

BY

P. H. ALSBIRK

In a material of 839 Greenland Eskimos with no corneal abnormalities, the variations of central corneal thickness (CT) were analysed from a genetic point of view. Environmental factors were suggested by a town-village difference in mean CT level (excess in town), by a higher CT level in families with an indoor occupation, and by a nearly significant husband-wife CT correlation. Using an age-sex-location-independent CT deviation score (CTDS) the family variations were studied in 86 families with two parents and one to seven children, and in 187 sibships with two to seven sibs. These first degree relationships showed a fairly high level of resemblance, which seems to indicate a major genetic influence on CT, with heritability esti- mates about 0.64.7. In men an association was found between CT and Caucasian admixture, which also suggested genetic factors. However, both findings call for reservation because of the socioeconomic environmental influence.

Key words: corneal thickness - population study - socioeconomic factors - genedenvironment - heritability - polygenic inheritance - Greenland Eskimos.

Variability of central corneal thickness (CT) in normal eyes is fairly pronounced, but the sources of this variation a r e not well known. In a recent population study a corneal thinning with increasing age was observed, especially pro-

Received October 5, 1977.

105

P . H . Alsbirk

nounced in men (Alsbirk 1978). Thus environmental influence was strongly suggested, possibly related to arctic conditions. The study was performed in Greenland Eskimos at 71' northern latitude. A similar reduction of C T with age was observed by Forsius et al. (1971) in Finns, Lapps and Skolts. The same group (Forsius et al. 1967) found a pronounced familial resemblance between offspring and parents, suggesting genetic influence in CT, but no regression Analyses and no husband-wife comparison were described. Apart from this study, familial correlations of CT have not been convincingly demonstrated before. Nakajima et al. (1968) found no significant heritability of CT.

In the present paper, some socioeconomic and familial variations in C T are analysed as part of an oculometric and genetic population study, based on 839 persons.

Material and Methods

The basic material of this study was 931 Eskimos, 98 O/o of the population above 7 years in the town and above 40 years in 7 villages of Umanaq District in West Greenland. A total of 839 persons was collected without corneal disease, scarring or anterior chamber deformity. Both eyes were measured in these persons who formed the material of a recent study of age-sex variations and oculometric correlations in CT (Alsbirk 1978). Table I shows the family groups available within this material. Consanguineous marital pairs (n = 22) as closely related as second cousins or closer, have been excluded from the table. The pedigrees had been registered along with the anterior chamber depth survey (Alsbirk 1975a), from which the present measurements are also derived.

Due to a significant CT decrease with age and a thinner cornea in male adults (Alsbirk 1978), the CT data available had to be transformed before the genetic analyses. Furthermore, a significant CT difference between the town and village groups (cf.

Table I . Family material available for child-parent and sib-sib correlation study of

corneal thickness (CT)

No. of children per family 1 2 3 4 5 6 7 Family group:

Mothers, fathers and offspring Total no. of children

Sibships Total no. of sibs

28 18 15 14 7 2 2 86

28 36 45 56 35 12 14 226

- 98 46 21 15 5 2 187

196 138 84 75 30 14 537

106

Corneal Thickness. I I . Heritability

Table 11) was found. Therefore, a standardized corneal thickness deviation score, CTDS, was calculated for every person, based on sex specific linear regressions of CT on age, in town and village subgroups. The positive or negative CT deviation of each individual was taken using the standard deviation about the appropriate regression ( s ~ , ~ ) as a unit of measurement. Following this transformation the pooled sample of 839 persons showed the expected distribution with mean 0.00, SD 1.00 and SEM 0.03. Following this transformation the environmental and genetic influence upon the residual CT variation could be analysed.

Linear regression studies and analyses of variances were carried out by means of standard computer programs. Heritability estimates were derived using standard pro- cedures of quantitative genetics, cf. Cavalli Sforza & Bodmer (1971) and Smith (1975). An analogous heritability study of anterior chamber depth has recently been described (Alsbirk 1975a,c).

Resu I ts

Environmental fac tors were suggested by the results given in Tables 11-IV, which show the variations of C T according to location, male occupation, and to CT of female spouse. On the average, 0.015 mm thicker corneas werc found in town dwellers of both sexes, compared to village inhabitants of the same age groups, cf. Table 11. Male indoor workers showed significantly higher C T values than the general population. However, no simple association with climatic factors seems to account for this finding, as the wives of these men showed the most pronounced deviations from expectation (= O.O), cf. Table 111. The total number of unrelated marital pairs are shown in Table IV. The elderly town families showed a significantly positive marital resemblance, village pairs did not.

Table I I . Corneal thickness in town and village populations of Umanaq District,

in age groups over 40 years.

Males

I mean I SD

No. of persons

Town Villages

t-test

63 0.515 0.029 110 0.500 0.032

P < 0.01

74 119

Females

C T (mm)

mean 1 SD

0.533 0.028 0.5 17 0.029

P < 0.001

107

P . H . Alsbirk

Males

No. 1 meanCTDS Male occupation

Table 111. Corneal thickness, in age-sex-location-independent deviation score units (CTDS),

according to male occupation, with their wives for comparison.

Females

No. mean CTDS

Location

'" Significance when compared to population mean value 0.00: P < 0.01. 1 CTDS unit - 0.029 mm.

Wife age No. of pairs b 1 sb 1 Significance

Genetic factors were gauged by the analyses given in Tables V-VII. The child on parent regression coefficients were fairly high, as an 0.5 value in childhingle parent and a 1 .O value in child/midparent relationships would have been expected on a purely additive genetic hypothesis. Probably, however, the estimates were to some extent inflated by environmental influence, as demon- strated by Table IV and the more relevant, pronounced marital resemblance within the families of Table V (b = 0.28, P < 0.05). The material was analysed in various subgroups of town and village families. The child on midparent

Table IV. Familial environmental influence on corneal thickness, estimated by husband on wife

regression coefficients (b).

Town - I 39 62 0.10 0.15 N. S. Town 40+ 33 0.32 0.12 P < 0.01 Villages 40+ 59 0.01 0.12 N. S.

Whole district 40+ 92 0.15 0.09 N. S. Whole district all 154 0.13 0.08 N. S.

s,, = standard error of b.

108

Corneal Thickness. 11. Heritability

Relatives (No.) I Father (86) Mother (86) Midparent (86)

Son (106) 0.35 0.11 0.45 0.09 0.59 0.1 1 Daughter (120) 0.47 0.1 1 0.38 0.09 0.62 0.12

~~ ~

Brothers

MS 1 d. f. Source of variation -

Child (226) 0.42 0.08 0.41 0.06 0.61 0.08

Sisters Sibs

MS 1 d. f. 7

Father (86) 0.28 0.11

regressions (k sb) were found to be 0.84 (0.19) in (town) families with husband age below 40 years against 0.56 (0.09) in (all) families with husband age above 40 years. However, this difference was not significant and the pooled material is given in Table V.

The sibs were analysed as Table VI shows. The intrapair correlation coeffi- cients (TI) indicated a level of resemblance within sibships similar to the re- semblance of offspring and single parents. In the sib subgroups aged less than 20 years, 20-39 years and above 40 years (by age of elder sib) the following

Table VI . CT in sibs. Analysis of variance and intracloss correlation coefficients (q).

F ratio rI k SE

2.40" 2.16" 2.45" 0.37 f 0.07 0.33 f 0.07 0.34 k 0.04

MS: mean square: d. f.: degrees of freedom. '$ P < 0.001

109

P . H . Alsbirk

Eskimo Sex

No. 1 meanCTDS

Table VII . Corneal thickness (in CTDS units) according to a classification of physiognomy

in 839 Greenlanders.

Significance t-test mixed

No. meanCTDS

1 CTDS unit - 0.029 mm.

rI values (k SE) were found 0.28 (0.07), 0.29 (0.06) and 0.44 (0.06) respectively. Thus no significant age variation was found. My working hypothesis: a greater resemblance in younger sibships, mostly living together, than in elderly, sepa- rated sibs, received no support, as in fact the opposite trend was found.

A coarse anthropological apfiroach was used also in the C T study, cf. Alsbirk (1975~). In male Greenlanders a significantly thicker cornea was found in those with a ‘mixed’, compared to those with an ‘Eskimoan’ facial appearance (Table VII, excess 0.010 mm). This finding closely corresponds to the pro- nounced ethnic difference in C T between male Danes and Greenlanders (cf. Alsbirk 1978, Fig. 3).

Discussion

Normal central corneal thickness (CT) ranges from about 0.43 to about 0.63 mm. The preceding paper showed that one source of this variation is related to age (Alsbirk 1978). In men the age effect amounted to 17 O/O of the total C T variation, in women less, about 3 O / o . This corneal thinning with age in Eskimos may be an environmental effect, perhaps a ‘secular trend’, but further studies in other populations are needed.

Within the present material certain environmental factors were analysed. Thus a remarkable socioeconomic variation was found. The comparison between elderly town and village inhabitants showed a higher C T average in the town (Table 11). A great difference in living circumstances still exists between these

110

Corneal Thickness. I I . Heritability

groups. The seven small village communities are isolated in the arctic fiord complex and represent a traditional economy based on hunting and fishing. The town population is mainly occupied in service and administration.

The town-village difference was statistically eliminated in the remaining analyses. Nevertheless, an interesting pattern appeared when (male) indoor workers (teachers, clerks, workmen and the like) were compared with hunters and fishermen. Although the two subgroups did not differ significantly from each other, the indoor workers had a thicker cornea than the average of the general population (CTDS 0.0, cf. Table 111). But, surprisingly, their wives had thicker corneas too. Thus my climatic working hypothesis was seriously attacked, as all women in the district spend most of their time indoors. There- fore, housing conditions, nutritional and general health factors may also play their part. Such socioeconomic variables must be taken into account in future corneal thickness surveys. A C T study of Greenland Eskimos who have spent many years in Denmark is being planned. Such a study might throw further light upon the socioeconomic C T variability.

Familial resemblance with respect to corneal thickness was found to be fairly great. The residual C T variation showed a high degree of parent-offspring and sib-sib similarity (Tables V and VI). The child on midparent regression coefficient was 0.61 (0.08), and doubling the intrapair correlation coefficient of sibs gave 0.68 (0.08). Thus, taken at face value, these estimates (with standard errors) indicated a heritability of about 0.6-0.7, i. e. about two thirds of the residual CT variation seemed to be genetically determined. This finding agrees with the heritability of many other physical measurements (Smith 1975) in- cluding results on corneal diameter, radius of curvature, anterior chamber depth and axial length (Alsbirk 1975b,c 1977). However, contrary to these ocular parameters, studied in the same population, an influence of the common familial environment seems to exist with respect to CT, possibly inflating the familial correlations to some extent. Thus the 86 parents of Table V showed a husband- wife similarity, which might be due to environmental influence, as assortative mating seems unlikely. Children resembled their parents slightly more in younger than in older families, but, on the other hand, sibs tended to be mutually most alike above the age of forty.

Thus no clear pattern emerged from the genetic studies. As the fairly large standard errors indicate, such studies in the field of quantitative genetics call for very large materials, if precise estimates of familial resemblance and heri- tability are wanted, not less if socioeconomic stratification is necessary, as the present findings seem to show. Furthermore, the previous methodological study showed that error of measurement is a problem, as 17 O/O of the total CT vari- ation seemed to be due to this type of ‘noise’ (Alsbirk 1974). Probably, however,

111

P . H . Alsbirk

the error of measurement should only be able to increase the non-genetic part of the phenotypic variation, i. e. it will tend relatively to lower its genetic proportion, the heritability estimate (cf. Smith 1975).

The demonstration of a thicker cornea in men of ‘mixed’ appearance compared to men of Eskimoan facial type is clear. Again genetic factors, with a certain socioeconomic contribution, might explain this result. The living circumstances differed enormously between Greenlanders and Danes in this area at the time of the survey, in 1969, but smaller variations among the Greenlanders associated with their degree of Caucasian admixture were also found. Unfortunately, the classification of facial types is far too coarse to permit further analyses along these lines (Alsbirk 1975~) .

As a conclusion: the population study of central corneal thickness (CT) among 839 Greenland Eskimos showed a resemblance between first degree re- latives, which seems to indicate a fairly high degree of genetic influence upon this parameter. However, influence of socioeconomic factors was also found, related to age, location and common familial environment. Due to the last factor the high level of heritability obtained (hy = 0.6-0.7) must be taken with reservation.

Acknowledgments

This and the preceding paper were supported by grants from Fabrikant Einar Willum- sens Mindelegat, the Danish Committee for Prevention of Blindness, the Danish Medical Research Council, the Commission for Scientific Research in Greenland, the Danish Hospital Foundation for Medical Research, Region of Copenhagen, the Faroe Islands and Greenland, and Carla Cornelia Storch Mellers legat.

References

Alsbirk P. H. (1974) Optical pachymetry of the anterior chamber. A methodological study of errors of measurement using Haag Streit 900 instruments. Acta o1dzthal.

Alsbirk P. H. (1975a) Anterior chamber depth and primary angle-closure glaucoma. 11. A genetic study. Acta ofihthal. (Khh.) 53, 436-449.

Alsbirk P. H. (1975b) Corneal diameter in Greenland Eskimos. Anthropometric and genetic studies with special reference to primary angle-closure glaucoma. Acta ~ p h -

Alsbirk P. H. (1975~) Anterior chamber of the eye. A genetic and anthropological study

Alsbirk P. H. (1977) Variation and heritability of ocular dimensions. A population study

(Kbh.) 52, 747-758.

t h d . (Kbh.) 53, 635-646.

in Greenland Eskimos. Human Heredity 23. 418-427.

among adult Greenland Eskimos. Acta ofihthal. (Kbh.) 55, 443-456.

112

Corneal Thickness. 11. Heritability

Alsbirk P. H. (1978) Corneal thickness. I. Age variation, sex difference, and oculometric correlations. Acta ophthal. (Kbh.) .56, 95-104.

Cavalli-Sforza L. L. & Bodmer W. (1971) The genetics of human populations, pp. 508- 633. Freeman, San Francisco.

Forsius H., Luukka H., Fellman J. & Eriksson A. W. (1967) Corneal thickness and its heredity in the population in North Finland. Bull. Eur. SOC. hum. Genet. I , 81-83.

Forsius H., Luukka H. & Eriksson A. W. (1971) Ophthalmogenetic studies on the Skolt Lapps. Acta ophthal. (Kbh.) 49, 498-502.

Nakajima A., Kimura T., Kitamura K., Uesugi M. & Handa Y. (1968) Studies on the heritability of some metric traits of the eye and the body. Jab. /. hum. Genet. 13,

Smith C. (1975) Quantitative inheritance. In: Fraser G. R. and Mayo O., Eds. Textbook 20-39.

of human genetics, pp. 382-441. Blackwell, Oxford.

Author’s address: P. H. Alsbirk, M. D., Granholmen 26, DK-2840 Holte, Denmark.

Acta ophthal. 56, I

113 8