3.2. Repeated measures ANOVA at 1 and 2 years of follow-up respectively:
Concerning the comparison of UDVA scores, Mauchly’s test indicated that the assumption of sphericity has not been met, χ2 (2) = 299.689, p < .001. Therefore degrees of freedom were corrected using Greenhouse-Geisser estimates of sphericity (ε = .517). A significant difference was found between the scores at baseline, 1 year and 2 years respectively, F(2, 224) = 12.71, p < .001, η2 = .102. Post-hoc pairwise comparisons showed a statistical significance of the scores when taken two by two (baseline vs 1 year 1.072 vs .074, p < .001; baseline vs 2 years 1.072 vs 0.086, p < .001, 1 year vs 2 years .074 vs .086, p = .043).
Concerning the comparison of CDVA scores, no significant difference was found between the scores at baseline, 1 year and 2 years respectively, F(2, 224) = 1.82, p = .165, η2 = .016. Post-hoc pairwise comparisons showed a statistical significance of the scores when taken two by two (baseline vs 1 year .009 vs .008, p = .957; baseline vs 2 years .009 vs .008, p = .957, 1 year vs 2 years .008 vs .008, p = 1).
Concerning the comparison of SPH scores, Mauchly’s test indicated that the assumption of sphericity has not been met, χ2 (2) = 534.595, p < .001. Therefore degrees of freedom were corrected using Greenhouse-Geisser estimates of sphericity (ε = .502). A significant difference was found between the scores at baseline, 1 year and 2 years respectively, F(2, 224) = 9.81, p < .001, η2 = .081. Post-hoc pairwise comparisons showed a statistical significance of the scores when taken two by two (baseline vs 1 year − 10.192 vs − .050, p < .001; baseline vs 2 years − 10.192 vs − .050, p < .001, 1 year vs 2 years − .050 vs − .071, p = .810).
Concerning the comparison of CYL scores, Mauchly’s test indicated that the assumption of sphericity has not been met, χ2 (2) = 148.201, p < .001. Therefore degrees of freedom were corrected using Greenhouse-Geisser estimates of sphericity (ε = .576). A significant difference was found between the scores at baseline, 1 year and 2 years respectively, F(2, 224) = .572, p = .565, η2 = .005. Post-hoc pairwise comparisons showed a statistical significance of the scores when taken two by two (baseline vs 1 year .881 vs .183, p < .001; baseline vs 2 years .881 vs .224, p < .001, 1 year vs 2 years .183 vs .224, p = .044).
Concerning the comparison of SPH EQ scores, Mauchly’s test indicated that the assumption of sphericity has not been met, χ2 (2) = 520.072, p < .001. Therefore degrees of freedom were corrected using Greenhouse-Geisser estimates of sphericity (ε = .502). A significant difference was found between the scores at baseline, 1 year and 2 years respectively, F(2, 224) = .572, p = .565, η2 = .005. Post-hoc pairwise comparisons showed a statistical significance of the scores when taken two by two (baseline vs 1 year 9.755 vs .041, p < .001; baseline vs 2 years 9.755 vs .041, p < .001, 1 year vs 2 years .041 vs .041, p = 1).
Concerning the comparison of IOP scores, Mauchly’s test indicated that the assumption of sphericity has been met, χ2 (2) = .979, p = .305. A significant difference was found between the scores at baseline, 1 year and 2 years respectively, F(2, 224) = .572, p = .565, η2 = .005. Post-hoc pairwise comparisons showed no statistical significance of the scores when taken two by two (baseline vs 1 year 12.353 vs 12.353, p = 1; baseline vs 2 years 12.353 vs 12.284, p = 1, 1 year vs 2 years 12.353 vs 12.284, p = 1).
Concerning the comparison of ECC scores, Mauchly’s test indicated that the assumption of sphericity has not been met, χ2 (2) = 929.29, p < .001. Therefore degrees of freedom were corrected using Greenhouse-Geisser estimates of sphericity (ε = .500). A significant difference was found between the scores at baseline, 1 year and 2 years respectively, F(2, 224) = .572, p = .565, η2 = .005. Post-hoc pairwise comparisons showed no statistical significance of the scores when taken two by two (baseline vs 1 year 2781.379 vs 2936.293, p = 1; baseline vs 2 years 2781.379 vs 2720.543, p < .001, 1 year vs 2 years 2936.293 vs 2720.543, p = .866).
It is of note that females had higher UDVA at 1 year (Beta = .048; p = .033), and higher CDVA at 1 year (Beta = .040; p = .049) and 2 years (Beta = .040; p = .049) compared to males respectively (Table 2).
Table 2
Repeated measures ANOVA of factors associated with the dependent variables at 1 and 2 years respectively.
Variable | Beta | p | 95% Cl | Partial Eta Squared |
Model 1a: UDVA at 1 year as the dependent variable |
Age | .001 | .695 | − .004 | .006 | .001 |
Gender (males vs females*) | .048 | .033 | .004 | .092 | .040 |
ICL type (toric vs non toric*) | .019 | .383 | − .025 | .063 | .007 |
Model 1b: UDVA at 2 years as the dependent variable |
Age | − .001 | .767 | − .005 | .004 | .001 |
Gender (males vs females*) | .033 | .135 | − .010 | .077 | .020 |
ICL type (toric vs non toric*) | .025 | .247 | − .018 | .069 | .012 |
Model 2a: CDVA at 1 year as the dependent variable |
Age | − .001 | .783 | − .005 | .004 | .001 |
Gender (males vs females*) | .040 | .049 | .001 | .081 | .034 |
ICL type (toric vs non toric*) | .022 | .282 | − .018 | .062 | .010 |
Model 2b: CDVA at 2 years as the dependent variable |
Age | − .001 | .783 | − .005 | .004 | .001 |
Gender (males vs females*) | .040 | .049 | .001 | .081 | .034 |
ICL type (toric vs non toric*) | .022 | .282 | − .018 | .062 | .010 |
Model 3a: SPH at 1 year as the dependent variable |
Age | − .004 | .552 | − .018 | .010 | .003 |
Gender (males vs females*) | − .042 | .513 | − .169 | .085 | .004 |
ICL type (toric vs non toric*) | − .012 | .850 | − .139 | .114 | < .001 |
Model 3b: SPH at 2 years as the dependent variable |
Age | − .001 | .870 | − .017 | .014 | < .001 |
Gender (males vs females*) | − .046 | .524 | − .190 | .097 | .004 |
ICL type (toric vs non toric*) | .011 | .879 | − .132 | .154 | < .001 |
Model 4a: CYL at 1 year as the dependent variable |
Age | .002 | .802 | − .011 | .014 | .001 |
Gender (males vs females*) | .069 | .226 | − .043 | .181 | .013 |
ICL type (toric vs non toric*) | .043 | .447 | − .069 | .155 | .005 |
Model 4b: CYL at 2 years as the dependent variable |
Age | .002 | .698 | − .010 | .014 | .001 |
Gender (males vs females*) | .031 | .573 | − .079 | .142 | .003 |
ICL type (toric vs non toric*) | .068 | .218 | − .041 | .178 | .014 |
Model 5a: SPH EQ at 1 year as the dependent variable |
Age | − .003 | .639 | − .018 | .011 | .002 |
Gender (males vs females*) | − .006 | .933 | − .139 | .127 | < .001 |
ICL type (toric vs non toric*) | .008 | .904 | − .124 | .140 | < .001 |
Model 5b: SPH EQ at 2 years as the dependent variable |
Age | < .001 | .988 | − .016 | .015 | < .001 |
Gender (males vs females*) | − .031 | .675 | − .175 | .114 | .002 |
ICL type (toric vs non toric*) | .045 | .533 | − .098 | .189 | .003 |
Model 6a: IOP at 1 year as the dependent variable |
Age | .044 | .368 | − .052 | .140 | .007 |
Gender (males vs females*) | − .134 | .767 | -1.029 | .761 | .001 |
ICL type (toric vs non toric*) | .125 | .782 | − .765 | 1.015 | .001 |
Model 6b: IOP at 2 years as the dependent variable |
Age | .035 | .443 | − .055 | .125 | .005 |
Gender (males vs females*) | − .005 | .990 | − .845 | .834 | < .001 |
ICL type (toric vs non toric*) | − .208 | .622 | -1.043 | .626 | .002 |
Model 7a: ECC at 1 year as the dependent variable |
Age | -21.359 | .640 | -111.628 | 68.909 | .002 |
Gender (males vs females*) | 451.269 | .290 | -389.115 | 1291.653 | .010 |
ICL type (toric vs non toric*) | 470.678 | .267 | -365.085 | 1306.441 | .011 |
Model 7b: ECC at 2 years as the dependent variable |
Age | 2.454 | .661 | -8.608 | 13.517 | .002 |
Gender (males vs females*) | 16.588 | .750 | -86.404 | 119.580 | .001 |
ICL type (toric vs non toric*) | 16.72 | .747 | -85.706 | 119.146 | .001 |
Numbers in bold indicate significant p values.