3. Results
From the total of 220 patients, 132 were males (60%) and 88 females (40%), where 181 had exophoria (82.3%) and 39 presented esophoria (17.7%). 91 patients were hyperopic (41.4%) and 21 myopic (9.6%) with or without astigmatism, 11 patients had pure astigmatism (5%), whereas 97 patients did not have any refractive error (44%). 214 participants had flat fusion (evaluated using the Worth dot test) at distance fixation (97.3%) and only 6 patients presented diplopia (2.7%), while at near fixation, flat fusion was presented in 199 participants (90.5%), 20 had diplopia (9%) and only 1 child showed suppression of one eye (0.5%).
1. Descriptive statistics and analysis based on gender.
Gender did not relate to the visual acuity or stereopsis. Nor the amount of phorias measured through the alternate Cover test, Howell test and Thorington test at near and distance fixation showed statistically significant differences between male and female participants. Differences on the refractive status of patients were only significant for the hyperopia, being
girls more hyperopic than boys. The independent sample t-Test was used to analyze differences between groups based on gender, but for the esophoria state, myopia, and astigmatism, where the analysis was performed using the Mann-Whitney test considering the small sample and the non-normal data distribution analyzed by the Shapiro-Wilk test.
Table 1 illustrates the mean value, standard deviation (Std), and p-value of the analyzed variables based on gender. The break and recovery of the NPC is not included in this table as it is analyzed separately.
2. Comparison of the magnitude of phorias at near and distance,
measured through the alternate Cover test, Howell test and Thorington test.
The Paired-Samples T-test was performed to analyze differences between the obtained measurements through the alternate Cover test, Howell test and Thorington test for XF, whereas, Wilcoxon for EF, considering the data distribution analyzed by Shapiro-Wilk. The alternate Cover test was compared to Howell test and posteriorly to Thorington test. On a second phase, the Howell test was compared to Thorington test.
Table 2 presents the results of the statistical analysis. Statistically significant differences were obtained from each test used, as seen in
Table 2, but for the esophoria state at far when measured through the Howell test and Thorington test. The highest phoria values were obtained by the alternate Cover test, whereas the lowest ones, by the Howell test, being these differences more noticeable at near fixation. The biggest difference was found on the magnitude of exophoria measured at near fixation by the alternate Cover test and Howell test. The minor difference, however, was seen on the magnitude of esophoria at distance fixation as defined by the alternate Cover test and Howell test.
3. Correlations among the phoria tests used to determine the most similar ones.
To find out, which tests have more similarities between them, and which one is the best to use at which distance or under which phoria conditions, the correlation analysis was performed. To statistically determine the correlations on the phoria values obtained at near and distance fixation through the alternate Cover test, Thorington test and Howell test, Pearson and Spearman correlation analysis was carried out, based on the normality of data distribution as confirmed by Shapiro-Wilk analysis. Refer to
Table 3(a,b).
Despite their different way of dissociation to determine the phoria state at distance and near fixation, some important correlations were found between them. In patients with exophoria, at near, similar values can be obtained by Thorington test and alternate Cover test, whereas at far, by Howell test and alternate Cover test. Thorington test and Howell test, however, show bigger differences on the obtained values. The same pattern was found in patients with esophoria, where the alternate Cover test and Thorington test relate to each other more significantly than Howell test and Thorington test. However, Thorington test and Howell test, present a strong correlation when quantifying the magnitude of esophoria at near fixation. Scatterplots to represent these correlations were created only for the phoria state at near, being the most important one, as it relates to the stereopsis degree. See
Figure 1,
Figure 2,
Figure 3 and
Figure 4.
4. Regression analysis to determine variables that could predict the degree of
stereopsis.
The regression analysis was performed to determine which variables could predict the degree of stereopsis. A new variable was introduced in this phase, the break of NPC, being this one related to the vergence state of patients. From the performed analysis, only the exophoria state at near and the break of NPC were related to the stereopsis value. The esophoric state was not related to the degree of stereopsis. The predictive models for every phoria test used and the break of NPC are presented in
Table 4.
Table 4 shows that the strongest predictive variable for the stereopsis degree is the break of NPC (higher standardized beta-value and F-value than the rest), followed by the near exophoria state measured through the Howell test, alternate Cover test and Thorington test as the last one.
Figure 5 depicts the correlation found between the stereopsis degree and the break of NPC, while
Figure 6 illustrates the correlation found between the stereopsis and the exophoria at near when the Howell test was used.
The relationship between the break of NPC and its recovery was also analyzed, where the Pearson correlation coefficient was of .96 and R
= .916. This means that break and recovery are proportional to each other, as it is expected to be.
Figure 7 shows this correlation.
5. Statistical analysis of the stereopsis degree, phoria state and the break of NPC
based on age.
Considering that the visual system changes during childhood, age is considered an important variable related to its performance. Therefore, three age-groups were created to analyze the phoria state, stereopsis, and break of NPC, as follows: 1
st group 5.06 to 8.00 (62 patients, 28.2%); 2
nd group 8.01 to 11.00 (111 patients, 50.5%), and 3
rd group 11.01 to 17.11 (47 patients, 21.4%). The One-Way Anova Test was used to analyze and compare the means among the three groups. The Levene test of homogeneity of variances showed p-values greater than .05, suggesting that variances among groups, were not significantly different. The Tuckey HSD was used for multiple comparison. Results from the analysis of stereopsis and break of NPC are presented in
Table 5.
Figure 8 and
Figure 9, represent the mean value and Std for the stereopsis degree and the break value of NPC as a dependent variable of age.
When it comes to the phoria state, age was not related to its magnitude. More specifically, taking the alternate Cover test as a reference (the most used technique in the clinical practice of optometrists and ophthalmologists to evaluate the phoria state), p =.19 and .11 for the XF value at distance and near fixation, whereas for EF, p = .51 and .92 for distance and near respectively.
6. Statistical analysis of the stereopsis degree and phoria state based on the
break value of NPC.
The break of NPC is associated with the exophoria state and the vergence system of an individual. Likewise, the exophoria can have an impact on the binocular state of the participant, like the stereopsis degree. Consequently, the break of NPC was divided into three groups as follows: the 1st group included all participants with a break from 0-5 cm (122 patients, 55.5 %); the 2nd group was composed by participants with a break from 6-10 cm (55 patients, 25 %), and the 3rd one, had a break from 11-25 cm (43 patients, 19.5 %). The phoria state and stereopsis degree was analyzed based on the break value of the NPC. Break values within the 10 cm range, are considered the norm.
As it can be seen by
Table 6, significant differences were found when the third group (break of NPC 11-25 cm), was compared to the first and second one. When the break value does not exceed the 10 cm range, is considered within normal limits. Therefore, patients with a break beyond it, present diminished degree of stereopsis. No statistically significant differences were found when group one and two were compared, as both have a break value withing normal ranges.
Figure 10 illustrates the mean stereopsis degree as a dependent variable of the break value of NPC, where the third group (break of 11-25 cm) obtained the lowest value and the first one (break of 0-5 cm), the highest one.
A total of 181 participants presented exophoria at near and distance fixation. From the total, 91 participants (50.3%), formed the first group (break of 0-5 cm), 47 participants (26%) formed the second group (break of 6-10 cm), and 43 participants (23.7%), were included in the third group (break of 11-25 cm). Results from the One-Way Anova with Tuckey HSD for multiple comparisons analysis are illustrated at
Table 7. Patients with esophoria could not be analyzed considering the small sample, which divided into groups, could not be representative.
Figure 11 represents the mean value and Std of the exophoria state at near fixation only (as presenting stronger correlation), as a dependent variable of the break value of NPC. Higher break value (worse) is related to higher magnitude of exophoria (worse) at distance and near fixation.