keratoconus Disease severity and family history in

27 downloads 122 Views 141KB Size Report
Aug 11, 2010 - L Szczotka-Flynn, M Slaughter, T McMahon, et al. keratoconus. Disease severity and family history in http://bjo.bmj.com/content/92/8/1108.full.
Downloaded from bjo.bmj.com on August 11, 2010 - Published by group.bmj.com

Disease severity and family history in keratoconus L Szczotka-Flynn, M Slaughter, T McMahon, et al. Br J Ophthalmol 2008 92: 1108-1111

doi: 10.1136/bjo.2007.130294

Updated information and services can be found at: http://bjo.bmj.com/content/92/8/1108.full.html

These include:

References

This article cites 25 articles, 5 of which can be accessed free at: http://bjo.bmj.com/content/92/8/1108.full.html#ref-list-1

Email alerting service

Topic collections

Receive free email alerts when new articles cite this article. Sign up in the box at the top right corner of the online article.

Articles on similar topics can be found in the following collections Cornea (332 articles) Ocular surface (398 articles)

Notes

To order reprints of this article go to: http://bjo.bmj.com/cgi/reprintform

To subscribe to British Journal of Ophthalmology go to: http://bjo.bmj.com/subscriptions

Downloaded from bjo.bmj.com on August 11, 2010 - Published by group.bmj.com

Clinical science

Disease severity and family history in keratoconus L Szczotka-Flynn,1 M Slaughter,2 T McMahon,3 J Barr,4 T Edrington,5 B Fink,4 J Lass,1 M Belin,6 S K Iyengar,2 CLEK Study Group 1

Department of Ophthalmology and Visual Sciences, Case Western Reserve University and University Hospitals Case Medical Center, Cleveland, OH, USA; 2 Department of Epidemiology and Biostatistics, Case Western Reserve University and University Hospitals Case Medical Center, Cleveland, OH, USA; 3 Department of Ophthalmology & Visual Sciences University of Illinois-Chicago, Chicago, IL, USA; 4 The Ohio State University School of Optometry, Columbus, OH, USA; 5 Southern California College of Optometry, Fullerton, CA, USA; 6 Albany Medical College, Albany, NY, USA Correspondence to: Dr L Szczotka-Flynn, Department of Ophthalmology and Visual Sciences, Case Western Reserve University, 11100 Euclid Ave., Bolwell Bldg. Suite 3200, Cleveland, OH 44106, USA; loretta.szczotka@uhhospitals. org Presented in parts at the 2007 Global Keratoconus Congress (Las Vegas, NV) and the 2007 Annual ARVO meeting (Ft. Lauderdale, FL). Accepted 26 April 2008

1108

ABSTRACT Background/aims: To determine if disease severity is associated with a family history of keratoconus. Methods: Markers of disease severity in the CLEK Study cohort were assessed to determine if they could discriminate individuals with and without family history. Logistic regression was used to examine association between corneal scarring, average corneal power, flat and steep keratometry readings, and higher-order root mean square (RMS) wavefront error with family history. Results: In univariate analyses, none of the severity indices had any significant associations with family history; however, contact lens use, gender, and Caucasian race were found to be significant predictors. After controlling for these confounders, there were no significant associations between any severity indices and family history. Conclusions: Presence or absence of family history is not associated with more severe clinical disease, at least when each marker for severity is considered independently. The results of this analysis are important for genetic studies of keratoconus in that it will allow recruitment of keratoconus patients across all stages of disease severity because it does not influence familial aggregation.

Several studies have suggested that keratoconus (KC) is genetic1 2 with excess familial clustering in single families and twins being reported. Wang et al2 estimated the prevalence of KC in first-degree relatives of index cases to be 3.34%. This represents an increase of 15–67 times over the general population prevalence of 0.05–0.23%. Because the severity of the disease varies from asymptomatic forme fruste conditions to disabling scarring requiring corneal transplantation, the true incidence, prevalence and familial aggregation are difficult to ascertain. Specifically, in asymptomatic conditions, KC may only be detected using sophisticated instrumentation such as videokeratography. In fact, Wang et al estimated broad sense heritability of indices for KC using videokeratography to be 60%. Thus, intermediate traits as measured by videokeratography show that KC is highly heritable. KC has shown association with rare genetic syndromes, such as Woodhouse Sakati syndrome3 and Down syndrome4 further supporting the genetic hypothesis. In previous reports, between 6% and 16% of patients with keratoconus have familial aggregation.5–8 However, to date no studies systematically assess association of multiple markers of disease severity and family history. It is plausible that more severe disease may be found in patients with a greater genetic load evidenced by acute cases clustering in families. Prior to initiating genetic

studies of KC, it is important to assess whether disease severity influences familial aggregation. If so, recruitment efforts would be targeted towards severe keratoconus patients. The objective of this study was to utilise intermediate phenotypic markers to determine if a greater disease severity was associated with a family history of KC.

METHODS The Collaborative Longitudinal Evaluation of Keratoconus (CLEK) Study was a natural history study of 1209 patients diagnosed as having KC who were examined annually for 8 years.8 The data utilised in this study were from 1208 patients who completed the Year 7 CLEK study visit, or the last CLEK Study visit available. Intermediate markers of disease severity were used to determine if they could discriminate individuals with and without family history, defined by self-report. The indices include corneal scarring, average corneal power (ACP), flat keratometry (FlatK) reading, steep keratometry (SteepK) reading, and higherorder first corneal surface wavefront RMS error (HORMSE) Data for the clinical severity variables were drawn from the most severe eye, defined as the worst ACP reading, or worst steepK if ACP was missing. Once the worst eye was established, all data for that eye were used in subsequent analyses. Eyes that had undergone corneal surgery were excluded. All intermediate markers of disease severity, except for corneal scarring, were drawn from videokeratography. Corneal scarring was included as a biomicroscopy defined clinical feature of advanced KC. The scaling of scarring followed the CLEK protocol for ‘‘gestalt scarring,’’ a measure of total scarring observed in the central cornea.9 Scarring was classified as an ordinal variable ranging from 0 (no scarring) to 4 (severe) in 0.5 increments. Raw topographic files were sent to the CLEK Corneal Topography Reading Center (CTRC) at the University of Illinois at Chicago for videokeratographic analysis. Uniform simulated indices were then computed utilising the Ohio State University Corneal Topography Tool (OSUCTT).10 Initially continuous variables, they were stratified into ordinal variables as listed in table 1. ACP is the average dioptric power of all points within the central 3 mm corneal zone10 and has been shown to be the best simulated index for corneal power.11 Simulated SteepK and FlatK indices were calculated from the pair of corneal meridians 90u apart with the greatest difference in average power. The greater average is SteepK and the lesser average is FlatK. HORMSE was used to represent corneal asymmetry and irregularity as Br J Ophthalmol 2008;92:1108–1111. doi:10.1136/bjo.2007.130294

Downloaded from bjo.bmj.com on August 11, 2010 - Published by group.bmj.com

Clinical science higher-order optical aberrations. HORMSE is the third-order and higher RMS error through the 27th term. The RMS wavefront error was derived from raw corneal topography data via VOL-CT software (version 6.58, Sarver and Associates, Carbondale, IL). Univariate logistic regression was used to examine corneal scarring, ACP, SteepK, FlatK, and HORMSE to determine if any of these variables individually predicted family history. Additionally, demographic variables such as contact lens use, age, gender and race were evaluated. Contact lens use was defined as wearing any type of contact lens for any length of time; however, in the CLEK Study, over 98% of the contactlens-wearing patients wore rigid lenses. Stratified analyses (chisquare) and logistic regression were used to detect associations between age, gender, race, contact lens use and family history to assess potential confounding between family history and these intermediate markers. In a second, multivariable logistic regression model, demographic confounding variables were controlled, and markers of disease severity were assessed for association with family history. Interactions between severity variables and confounders were assessed in a separate model. The analyses were performed with the corneal curvature related-disease severity variables kept continuous as well as stratified as shown in table 1. Significance was set at the 0.05 alpha level. The Hosmer and Lemeshow goodness-of-fit test was used to assess the quality of the logistic regression models. Using a priori assumptions based on the current literature2 and final sample sizes of n = 203 and 937 (those that had corneal power data available and a positive or negative family history, respectively), we conducted a power analysis using a two-tailed test. We assumed a difference of 1.4 D difference in corneal power between those with a positive family history and those without, a standard deviation of 3.5 D in intrasubject difference and an alpha = 0.05. Under these assumptions, our sample had 99% power to detect differences in corneal power between those with and without family history.

RESULTS This report includes data on 1143 patients, since 65 had missing family history data. Almost 18% of patients (204 of 1143, 17.8%) reported a positive family history. Seventy per cent of the data were from Year 7 CLEK exams, and the remainder came from earlier exams. Interestingly, this percentage is higher Table 1 Categorisation of continuous variables Clinical indicators of severity

n

Mean (SD)

Corneal scarring RMS wavefront error Mild (RMS(3.5) RMS moderate (3.5,RMS (5.75) RMS severe (RMS.5.75) Average corneal power (dioptres) Mild (ACP(52) Moderate (52,ACP(56) Severe (ACP.56) Mean steep K (dioptres) Mild (steep K(45) Moderate (45,steep K(52) Severe (steep K.52) Mean flat K (dioptres) Mild (flat K(45) Moderate (45,flat K(52) Severe (flat K.52)

1085

0.786 (1.1)

531 191 70

2.11 (0.82) 4.36 (0.58) 9.45 (8.8)

524 241 286

48.1 (2.5) 54.0 (1.1) 63.3 (7.1)

53 555 597

43.9 (0.88) 49.0 (2.0) 59.0 (5.3)

252 585 368

43.2 (1.6) 48.2 (2.0) 58.3 (5.3)

Br J Ophthalmol 2008;92:1108–1111. doi:10.1136/bjo.2007.130294

Table 2 Unadjusted logistic regression results of disease variables and family history Clinical indicators of severity

Odds ratio (95% Wald CI)

p Value

Corneal scarring RMS wavefront error* RMS moderate RMS severe Average corneal power* Moderate Severe Mean steep K* Moderate Severe Mean flat K* Moderate Severe

1.03 (0.89 to 1.19)

0.70

1.11 (0.71 to 1.73) 1.05 (0.52 to 2.09)

0.65 0.90

1.13 (0.75 to 1.70) 1.01 (0.68 to 1.49)

0.56 0.98

1.26 (0.55 to 2.89) 1.37 (0.60 to 3.13)

0.59 0.46

0.82 (0.55 to 1.22) 0.95 (0.62 to 1.44)

0.32 0.79

n (% positive family history) 1028 (17.8%) 748 (17.2%)

996 (17.3%)

1140 (17.8%)

1140 (17.8%)

*Reference category: mild.

than previously reported in the CLEK Study where at the baseline visit, about 14% reported a family history.8 This change may reflect new incidence of disease, or a form of information bias due to increased disease awareness among family members over the 8 years of CLEK follow-up. Of the continuous intermediate disease marker variables included in the analysis, the means and standard deviations were as follows: HORMSE (3.3, ¡3.4), ACP (53.6 D, ¡7.6 D), FlatK (50.2 D, ¡6.6 D), SteepK (53.8 D, ¡6.7 D). Table 1 lists the sample size available for each intermediate marker and the means and standard deviations of each ordinal variable subclassification. No significant univariate associations between variables indicating disease severity and family history were observed (table 2). Continuous disease severity variables also showed no association (p>0.3572). Table 2 lists the sample size available for each univariate model and percentage of patients with a positive family history in that subsample. Despite missing data for some of the severity measures, the percentage of patients with a positive family history remained constant and consistent with the overall percentage of patients reporting a positive family history of disease. Stratified analyses of potential confounders found contact lens use (p = 0.0238), gender (p,0.0001) and race (p = 0.013) to be associated with a positive family history. Specifically, contact lens wearers, females and non-African Americans were associated with reporting a positive family history of KC. Within the race classifications, Caucasians had a significantly greater chance of reporting a family history compared with African Americans (OR 1.72, 95% CI 1.11, 2.7). Similar associations were found when adjusting for all confounders simultaneously (table 3). Table 3 Logistic regression results of potential confounders (n = 1118) Confounding variable

p Value

Odds ratio (95% Wald CI)

Age (reference ,25 years) 25–35 35–45 45–55 .55 Contact lens use Gender (reference females) Race (reference non-AA)

0.98 0.27 0.34 0.81 0.052 ,0.001* 0.003*

0.99 0.60 0.64 0.89 1.54 0.48 0.51

(0.39 (0.24 (0.26 (0.36 (0.99 (0.35 (0.33

to to to to to to to

2.51) 1.50) 1.59) 2.24) 2.38) 0.66) 0.80)

*Significant at the 0.05 alpha level.

1109

Downloaded from bjo.bmj.com on August 11, 2010 - Published by group.bmj.com

Clinical science Table 4 Adjusted logistic regression results of disease variables and family history Clinical indicators of severity

Odds ratio (95% Wald CI)

p Value

Corneal scarring RMS wavefront error* RMS moderate RMS severe Average corneal power* Moderate Severe Mean steep K* Moderate Severe Mean flat K* Moderate Severe

1.05 (0.90 to 1.22)

0.57

1.12 (0.71 to 1.77) 1.17 (0.57 to 2.38)

0.63 0.67

1.12 (0.74 to 1.71) 0.99 (0.66 to 1.49)

0.58 0.95

1.17 (0.47 to 2.89) 1.31 (0.53 to 3.24)

0.73 0.56

0.67 (0.44 to 1.02) 0.83 (0.53 to 1.30)

0.06 0.40

n (% positive family history) 1006 (18.0%) 732 (17.3%)

974 (17.4%)

1115 (17.8%)

1115 (17.8%)

*Reference category: mild.

Multivariable logistic regression results of disease marker variables and family history adjusted for contact lens use, gender and race are presented in table 4. None of the disease variables are significant. Similarly, when the corneal curvaturerelated disease variables were kept continuous, no associations with disease severity were detected. Table 4 also lists the sample size available for each multivariable model and percentage of patients with a positive family history in that subsample. Despite missing data for some severity measures, the percentage of patients with positive family history remained constant and consistent with the overall percentage of patients reporting a positive family history. As expected, sample sizes available for each model in table 4 are slightly smaller than the respective univariate models in table 2, due to a requirement for all measures to have non-missing values. We determined that sample sizes in table 4 are never less than 97.8% of the respective univariate models in table 2, and both results are likely based on similar groups of patients. Interactions between severity variables and confounders were assessed in a separate model. A global likelihood ratio test revealed no significant interactions (p = 0.103). The models show no evidence of lack of fit based on the Hosmer–Lemeshow statistics (p.0.29).

DISCUSSION Using intermediate phenotypic markers of disease progression, KC severity is not associated with family history of disease. Our findings are similar to others looking at family history and disease severity. Tuft et al have shown that family history of disease is not predictive of subsequent penetrating keratoplasty.12 Rabinowitz et al, have found corneal irregularity and asymmetry suggestive of variable forms of KC in family members of affected patients.7 The findings in family members were similar to, but less severe than, those found in KC patients, and may represent variable expression of genes contributing to KC development. Barr et al, using the CLEK Study sample but different methodology, did not find family history to be predictive of incident corneal scarring.9 Although disease severity is not associated with family history of disease in the CLEK Study, other analyses from the CLEK cohort revealed associations between disease severity and disease asymmetry or biomicroscopic findings. Specifically, CLEK Study patients with more severe disease were also more asymmetric in their disease status,13 and more advanced disease was associated with a greater likelihood of Vogt striae, Fleischer ring and/or corneal scarring.14 1110

Keratoconus disease severity can be defined by multiple methods. Traditionally, corneal curvature has been utilised with various stratification schemes. The stratifications we have chosen to represent mild, moderate and severe disease are common stratifications used by clinicians. Corneal scarring and HORMSE are two other methods to define disease severity. We have included corneal scarring as an intermediate marker of disease severity because it differentiates the more advanced disease states as is used in the Amsler–Krumeich and Keratoconus Severity Score (KSS) classification schemes.11 15 Finally, higher-order aberrations show potential to differentiate stages of KC disease severity. Eyes with KC have higher-order aberrations that are approximately 5.5 times greater than those found in normal eyes with vertical coma serving as the dominant type.16 17 Corneal aberrometry derived from topography has been argued as a better method to analyse the highly aberrated corneal surface in KC, as opposed to global ocular aberrometry which is limited to the pupillary area as in Hartmann–Shack sensor-based approaches.15 Higher-order RMS values were shown to discriminate early KC from normal eyes,18 and disease severity in KC.11 15 Therefore, we have used topographically derived corneal aberrometry as an intermediate marker of disease severity, as it is becoming a more common advanced approach to detecting and grading KC disease severity. In summary, selection of these intermediate markers was chosen based on classical clinical stratifications, clinical slitlamp signs that better differentiate more severe disease states, and sophisticated mathematical modelling of anterior corneal surface topography. Although these markers are distinct, many of them are somewhat correlated. For example, mean central keratometry is correlated with RMS of corneal coma-like aberrations15 and direct or indirect measurements of corneal power are associated with corneal scarring.9 19 20 Therefore, we would expect similar trends of association between family history and these markers of disease severity, as we have found. Keratoconus disease severity can also be summarised using a composite score in which a set of clinical signs and symptoms that vary collectively earmark compromised visual function. Examples of composite scores include classification schemes described by Rabinowitz21 with three categories: keratoconus, early keratoconus, and keratoconus suspect; and one described by Bennett which classifies keratoconus into four stages varying from Stage 1 which has such early disease that spectacles are the first form of treatment to Stage 4 which has corneal steepening .55 D, apical opacities and Munson’s sign.22 More recent composite severity scores include Alio’s recommended scheme utilising corneal higher order aberrations,15 and the KSS score developed by our group.11 The KSS score utilises the intermediate markers we have utilised in this study as well as visual corneal topographic shape analysis and presence of other KC slit-lamp signs. It was not utilised herein because all of the variables were not available on this cohort at the time of this study. Nevertheless, results may differ if a composite trait such as KSS were used. We adjusted for the effect of gender, race and contact lens use. In the stratified analysis and univariate logistic regression for gender, females were twice as likely to report a positive family history. This observation was previously published by the CLEK Study Group.23 Although it is possible that females have a greater predisposition to disease, this may simply be a reporting bias. Gender was also considered a confounder because previous reports describe differences in KC disease severity between males and females, when measured by visual acuity and certain slit-lamp signs.23 In the stratified and univariate Br J Ophthalmol 2008;92:1108–1111. doi:10.1136/bjo.2007.130294

Downloaded from bjo.bmj.com on August 11, 2010 - Published by group.bmj.com

Clinical science logistic regression for race, Caucasians were 72% more likely to have a self-reported family history as compared with African– Americans. This observation may be a reporting bias, or as described in the literature, differences in prevalence may reflect diversity in genetic susceptibility. For example, Asians, particularly of Pakistani origin, have been shown to have a higher prevalence, incidence and disease severity than white patients in the UK,24–26 and in New Zealand, topographic indicators of early KC were more prevalent in Maori/Polynesian patients compared with those of European descent.27 Lastly, contact lens use was a predictor of family history, and it has been associated with more severe disease.9 19 27 The associations between gender, race and contact lens use with both disease severity and selfreported family history led to these confounders being controlled for in the final multivariate analysis. A limitation of our study is that we did not examine the families of patients with mild, modest and severe disease directly. Careful examination may reveal that excessive clustering of disease, defined by more subtle features, occurs in families of patients with greater severity. Thus, not only may there be an under-reporting bias due to self-report, but there may be undetected early or form fruste KC in family members of these affected CLEK patients. In summary, the presence of family history was not associated with severe clinical outcomes for any of the severity variables examined. The results of this analysis permit recruitment from all KC patients in genetic studies regardless of diagnostic disease state because disease severity does not influence familial aggregation, bearing in mind that KC is likely oligogenic. Funding: Supported by National Eye Institute Grants R21 EY015145 (JL); EY10419, EY10069, EY10077, EY02687, EY12656 (CLEK); RO3EY17571 (TM); Research to Prevent Blindness (JL); American Optometric Foundation (LSF); and Ohio Lion’s Eye Research Foundation (LSF, JL).

4. 5. 6. 7. 8. 9.

10. 11.

12. 13. 14.

15. 16.

17.

18. 19.

20.

21.

Competing interests: Since the original submission of this manuscript, JB is now an employee of Bausch & Lomb; MB is a consultant to Oculus USA. Other authors: none.

22.

Ethics approval: Ethics approval was obtained.

23.

Patient consent: Obtained.

24.

REFERENCES 1. 2. 3.

Owens H, Gamble G. A profile of keratoconus in New Zealand. Cornea 2003;22:122–5. Wang Y, Rabinowitz YS, Rotter JI, et al. Genetic epidemiological study of keratoconus: evidence for major gene determination. Am J Med Genet 2000;93: 403–9. Al-Swailem SA, Al-Assiri AA, Al-Torbak AA. Woodhouse Sakati syndrome associated with bilateral keratoconus. Br J Ophthalmol 2006;90:116–17.

Br J Ophthalmol 2008;92:1108–1111. doi:10.1136/bjo.2007.130294

25.

26. 27.

Shapiro MB, France TD. The ocular features of Down’s syndrome. Am J Ophthalmol 1985;99:659–63. Assiri AA, Yousuf BI, Quantock AJ, et al. Incidence and severity of keratoconus in Asir Province, Saudi Arabia. Br J Ophthalmol 2005;89:1403–6. Lee LR, Readshaw G, Hirst LW. Keratoconus: the clinical experience of a Brisbane ophthalmologist. Ophthalmic Epidemiol 1996;3:119–25. Rabinowitz YS, Garbus J, McDonnell PJ. Computer-assisted corneal topography in family members of patients with keratoconus. Arch Ophthalmol 1990;108:365–71. Zadnik K, Barr J, Edrington T, et al. Baseline findings in the Collaborative Longitudinal Evaluation of Keratoconus (CLEK) study. Invest Ophthalmol Vis Sci 1998;39:2537–46. Barr JT, Wilson BS, Gordon MO, et al. Estimation of the incidence and factors predictive of corneal scarring in the Collaborative Longitudinal Evaluation of Keratoconus (CLEK) study. Cornea 2006;25:16–25. Mahmoud AM, Roberts C, Lembach R, et al. Simulation of machine-specific topographic indices for use across platforms. Optom Vis Sci 2006;83:682–93. McMahon TT, Szczotka-Flynn L, Barr JT, et al. A new method for grading the severity of keratoconus: the Keratoconus Severity Score (KSS). Cornea 2006;25:794–800. Tuft SJ, Moodaley LC, Gregory WM, et al. Prognostic factors for the progression of keratoconus. Ophthalmology 1994;101:439–47. Nichols JJ, Steger-May K, Edrington TB, et al. The relation between disease asymmetry and severity in keratoconus. Br J Ophthalmol 2004;88:788–91. Zadnik K, Barr JT, Gordon MO, et al. Biomicroscopic signs and disease severity in keratoconus. Collaborative Longitudinal Evaluation of Keratoconus (CLEK) study group. Cornea 1996;15:139–46. Alio JL, Shabayek MH. Corneal higher order aberrations: a method to grade keratoconus. J Refract Surg 2006;22:539–45. Maeda N, Fujikado T, Kuroda T, et al. Wavefront aberrations measured with Hartmann–Shack sensor in patients with keratoconus. Ophthalmology 2002;109:1996–2003. Pantanelli S, Macrae S, Jeong TM, et al. Characterizing the wave aberration in eyes with keratoconus or penetrating keratoplasty using a high-dynamic range wavefront sensor. Ophthalmology, 2007;114:2013–21. Buhren J, Kuhne C, Kohnen T. Defining subclinical keratoconus using corneal firstsurface higher-order aberrations. Am J Ophthalmol 2007;143:381–9. Barr J, Zadnik K, Wilson B, et al. Factors associated with corneal scarring in the Collaborative Longitudinal Evaluation of Keratoconus (CLEK) study. Cornea 2000;19:501–7. Weed KH, Macewen CJ, Giles T, et al. The Dundee University Scottish Keratoconus study: demographics, corneal signs, associated diseases, and eye rubbing. Eye 2008;22:534–41. Rabinowitz YS. Corneal topography. In: Bennet ES, Weissman BA, eds. Clinical contact lens practice. Philadelphia: Lippincott, Williams & Wilkins, 2005:215–32. Bennett E, Cutler S. Keratoconus. In: Bennett ES, Hom, MM, eds. Manual of gas permeable contact lenses. Philadelphia: Lippincott, 2004:237. Fink BA, Wagner H, Steger-May K, et al. Differences in keratoconus as a function of gender. Am J Ophthalmol 2005;140:459–68. Cozma I, Atherley C, James NJ. Influence of ethnic origin on the incidence of keratoconus and associated atopic disease in Asian and white patients. Eye 2005;19:924–5. [author reply 925–6] Georgiou T, Funnell CL, Cassels-Brown A, et al. Influence of ethnic origin on the incidence of keratoconus and associated atopic disease in Asians and white patients. Eye 2004;18:379–83. Pearson AR, Soneji B, Sarvananthan N, et al. Does ethnic origin influence the incidence or severity of keratoconus? Eye 2000;14:625–8. Owens H, Gamble GD, Bjornholdt MC, et al. Topographic indications of emerging keratoconus in teenage New Zealanders. Cornea 2007;26:312–18.

1111