Data was entered into a computer and analysed using Statistical Package for Social Sciences (SPSS) for windows (SPSS INC. USA) version 18.0. Categorical variables such as sociodemographic characteristics of respondents were presented on contingency tables for both groups, and these variables were compared between the glaucomatous and control groups using chi-square test. Continuous variables like LT and ONSD were presented by their means and standard deviations (mean ± SD) for both glaucomatous and control groups. LT and ONSD in both groups were compared statistically using independent sample t-test. Correlation and regression analysis were used to evaluate the relationship between age, LT and ONSD in the
A
Fig 6: B-mode sonographic measurement of the LT (axial plane), (A→B) seen as the anechoic space between the echogenic capsules between the cursors.
31
glaucomatous group while controlling for other variables.. The level of statistical significance was determined at p values less than 0.05 and the results were presented in frequency distribution tables, charts and graphs.
ETHICAL CONSIDERATIONS
The Approval for this study (Appendix 11) was sought and obtained from the Ethical and
Research committee of the Obafemi Awolowo University Teaching Hospital Complex IleIfe, Osun state, South-Western Nigeria.
32
RESULTS
Table I. Patients' General Characteristics
P Variables Glaucoma Controls χ2 df value
(n = 60) (n = 60) Age (years)
30 – 39 7 (11.7) 13 (21.7)
40 – 49 50 – 59
16 (26.7) 11 (18.3)
15 (25.0) 12 (20.0)
2.951 4 0.566
60 – 69 19 (31.7) 16 (26.7)
≥ 70 7 (11.7) 4 (6.7)
Mean ± SD 55.5 ± 11.9 52.1 ± 11.9 1.561 118 0.121*
(Min - Max) Gender
33.0 - 79.0 31.0 - 78.0
Male 32 (53.3) 27 (45.0)
0.834 1 0.361
Female 28 (46.7) 33 (55.0)
Weight (Mean ± SD) (Kg) 64.8 ± 11.2 70.3 ± 16.0 -2.320 118 0.033*
Height (Mean ± SD) (m) 1.65 ± 0.10 1.63 ± 0.08 0.717 118 0.475*
BMI (Mean ± SD) (Kg/m2) BMI categories
24.10 ± 4.83 26.21 ± 5.14 -2.320 118 0.022*
Underweight 6 (10.0) 4 (6.7)
Normal 33 (55.0) 25 (41.7)
3.673 3 0.299
Overweight 13 (21.7) 17 (28.3)
Patient category, n (%)
33
Obese
Duration of glaucoma (Mean ± SD)
8 (13.3) 14 (23.3)
(months) 1.2 ± 0.4 NA
χ2 - Chi square; df - degree of freedom; * independent samples t test; SD - Standard deviation
Table 1 summarises the age, sex and other characteristics of the participants of this study. In this prospective case control study, a total of 120 subjects between the ages of 30 and 80 years were included. They comprised 60 glaucoma subjects and 60 controls. Thirty-two (53.3%) of the glaucomatous subjects were males and 28 (46.7%) were females while there were 27(45%) males and 33(55%) females in the control subjects. This difference in their gender composition was however not statistically significant (x2=0.834, df=1, p=0.361). The mean age of the glaucoma subjects was 55±11.9 years while that of the control was 52.1±11.9 years (t = 1.561, df = 118, p = 0.121.). Participants in both study groups were also statistically similar in other general characteristics except for their weight and BMI. The mean weight and BMI for the control subjects were greater than those for glaucoma subjects, these values were statistically significant with p values of 0.033 and 0.022 for weight and BMI respectively. However more than half of the subjects investigated had normal weight with only a few being underweight, overweight and obese (x2 = 3.673, df = 3, p = 0.299) (Table 1). Of the 60 glaucomatous subjects, 54(90.0%) had POAG, 5(8.3%) had normal tension glaucoma and 1(1.7%) had angle closure glaucoma (Fig.7).
34
Fig. 7: Bar chart showing distribution of different types of glaucoma Table II. Differences in IOP, ONSD and LT between the study groups
Patient category
Variables Glaucoma Controls t df P value
(n = 60) (n = 60)
Right IOP (mmHg) 17.3 ± 8.1 14.5 ± 3.1 2.560 118 0.012 Left IOP (mmHg) 17.1 ± 7.6 15.0 ± 3.3 2.034 118 0.045
(90.0 %
54 )
1 (1.7 % ) 5 (8.3 % )
0 10 20 30 40 50 60
POAG Normal tension Glaucoma ACG
Type of Glaucoma
35
Right ONSD (mm) 3.57 ± 0.19 4.23 ± 0.34 -12.881 118 < 0.001 Left ONSD (mm) 3.59 ± 0.33 4.26 ± 0.30 -11.441 118 < 0.001 Right lens thickness (mm) 4.15 ± 0.43 4.01 ± 0.56 1.615 118 0.109 Left lens thickness (mm) 4.18 ± 0.46 3.99 ± 0.45 2.285 118 0.024
COMPARISON BETWEEN ONSD AND LT AMONG BOTH GROUPS
Table II shows the difference in, ONSD and LT between the glaucoma and control groups. The mean ONSD in glaucoma patients was 3.57 ± 0.19 mm on the right and 3.59 ± 0.33mm on the left. The mean ONSD among controls were 4.23 ± 0.34 mm and 4.26 ± 0.30mm on the right and left respectively. The right ONSD of the glaucomatous patients had a lower mean of
3.57 ± 0.19 mm compared to that of the controls of mean of 4.23 ± 0.34 mm. Their left mean ONSD was also lower (3.59 ± 0.33mm) than that of the controls (4.26 ± 0.30mm) (Table II).
36
These are also highlighted in figures 8 and 9 which showed statistically significant differences between both groups (p = 0.000).
Right LT of the glaucomatous patients had a mean of 4.15 ± 0.43mm while the controls had a mean of 4.01 ± 0.56mm (p = 0.109) (fig. 10). The mean left LT of the glaucomatous patients was however higher at 4.18 ± 0.46mm compared to that of the controls at 3.99 ± 0.45mm (fig 11). In addition the mean intraocular pressure of both eyes of glaucoma subjects was significantly higher than the mean for controls as shown on Table II.
37
Fig 8: Boxplot showing the pattern of Right ONSD among glaucomatous patients
(p = 0.000)
38
Fig 9: Boxplot showing the pattern of Left ONSD among glaucomatous
patients (p = 0.000)
39
Fig. 10: Boxplot showing pattern of right lens thickness among study
participants (p = 0.109)
40
Fig. 11: Boxplot showing pattern of left lens thickness among glaucoma patients (p = 0.024)
.
41
Table III. Paired comparison of IOP, ONSD and Lens thickness among study subjects
Mean ± SD
Variables t* P value
Right Left
IOP (mmHg)
Glaucoma 17.3 ± 8.1 17.1 ± 7.6 0.218 0.828
Controls ONSD (mm)
14.5 ± 3.1 15.0 ± 3.3 -2.259 0.028
Glaucoma 3.57 ± 0.19 3.59 ± 0.33 -0.513 0.610
Controls
Lens thickness (mm)
4.23 ± 0.34 4.26 ± 0.30 -1.413 0.163
Glaucoma 4.15 ± 0.43 4.18 ± 0.46 -0.967 0.338
Controls 4.01 ± 0.56 3.99 ± 0.45 0.341 0.734
* Paired-samples t test
42
As highlighted by a paired comparison of right and left eyes on Table III, there were no statistically significant differences in both ONSD and LT of glaucoma subjects and controls.
The mean ONSD values were 3.57 ± 0.19mm and 3.59 ± 0.33mm for glaucomatous eyes on the right and left (p = 0.610) while the control subjects had mean values of 4.23 ± 0.34mm and 4.26
± 0.3mm (p = 0.163) on the right and left respectively. Similarly, the mean LT were 4.15 ± 0.43mm and 4.18 ± 0.46mm for glaucomatous eyes on the right and left (p = 0.338) while the control subjects had mean values of 4.01 ± 0.56mm and 3.99 ± 0.45mm (p = 0.734) on the right and left respectively(Table III).
Table IV. Correlation between respondents' Age, ONSD and Lens thickness
43
Correlation – Age vs. :
Glaucoma Controls
R P value r P value
Right ONSD -0.230 0.077 -0.549 0.000
Left ONSD -0.191 0.144 -0.548 0.000
Right lens thickness 0.456 0.000 0.068 0.603
Left lens thickness 0.408 0.001 -0.032 0.811
r - Pearson correlation coefficient,
44
RELATIONSHIP BETWEEN AGE, ONSD AND LENS THICKNESS
A weak and negative correlation was observed between the age and ONSD of glaucoma patients in both eyes with Pearson correlation coefficients of -0.230 (p = 0.077) and -0.119 (p = 0.144) on the right and left respectively, while a strong and statistically significant negative correlation was observed in the control subjects (p = 0.000) on either sides (Table IV).
The LT strongly correlated with the age of the glaucoma patients on the right(r = 0.456, p = 0.000) and left (r = 0.408, p = 0.001) respectively. No significant correlation was observed between the LT and age of control subjects on the right and left sides (Table IV).
45
Table V. Simple linear Regression analyses for predictors of lens thickness in Glaucoma patients
Model Dependent variables
Independent variables
B SE t P value R2
1
Right lens thickness (RLT) (mm)
(Constant) 3.235 0.208
2
Left lens
thickness (LLT) (mm)
Age (years) 0.017 0.004 3.906 0.000
(Constant) Age (years)
3.314 0.016
0.005
3.402
0.001 0.166
Regression equations
Mode1: RLT (mm) = 3.235 + 0.017*Age (years) Mode2: LLT (mm) = 3.314 + 0.016*Age (years)
B - regression coefficients; SE - standard error; R2 - Model variance
46
As highlighted on Table V, bivariate analysis in glaucomatous patients showed weak but significant positive correlation between right and left LT with age, and a weak negative correlation with height for the left LT only (r = 0.456, p =0.000; r = 0.408, p = 0.001; r = 0.301, p = 0.019 respectively). Consequently, linear regression analyses were performed to ascertain the influence of age alone (Models 1 and 2) on the right and left LT as shown on
Table V. Age of the patients significantly predicted the respective lens thickness values in the Models. In Model 1, Age was responsible for only 20.8% (Model variance, R2 = 0.208) of the variation observed in right LT compared to 16.6% in Model 2 for the left LT.
47
Table VI. Discordant / Regular pattern of the IOP and ONSD
Variables Glaucoma Controls χ2 Df P value (n = 60) (n = 60)
Discordant 40 (66.7) 29 (48.3)
4.126 1 0.042 Regular pattern 20 (33.3) 31 (51.7)
RELATIONSHIP BETWEEN IOP AND ONSD
Patient category, n (%)
48
Although no statistically significant difference was observed between the mean IOP values of the right and left eyes of glaucomatous patients, the mean values of 14.1±0.31mm and 15.0±3.3mm on the right and left respectively among the controls showed a statistically significant difference(t = -2.259, p = 0.028) as shown on Table III. A comparison of IOP with ONSD showed a predominant discordant pattern of relationship among glaucomatous patients (discordant meaning an increase IOP was associated with an increase in ONSD) accounting for 66.7% of all cases, meanwhile only about a third of glaucoma patients showed the regular pattern (regular meaning an increase in IOP was related to a decreasing ONSD) (χ2 = 4.126, df
= 1; p = 0.042). Among the controls there was just a marginal increase of the regular pattern over the discordant pattern as shown on Table VI.
DISCUSSION
49
Glaucoma has consistently ranked the second leading cause of blindness10-12 and the leading cause of preventable and rather unfortunately, irreversible blindness globally11,13. The definition of this disease continues to evolve, allowing for continual modification of detection methods, treatment end-points and therapeutic options14. Glaucoma is traditionally assessed by the triad of tonometry, visual field testing and optic nerve evaluation15,16. Several ocular and sometimes non-ocular factors can however impair the usefulness of these morphological and functional assessment variables in the evaluation of glaucoma suspects15. These factors include but are not limited to an opaque media i.e. cataract, mental and sometimes physical limitations of patients15. There arises a need therefore to look away from this traditional triad of assessment to a more objective way of evaluating glaucoma, independently of IOP and optic nerve head morphology.
This present study set out to evaluate the ONSD with the aid of ultrasound. It was a prospective study carried out among Nigerian adults between the ages of 30 and 80 years. There was a male gender preponderance, a finding which was similar to other studies done in the past8,47,48. However a study conducted by Kosoko-Lasaki et al49 reported that this finding was inconsistent, this was corroborated by a study done in Ghana by Ntim-Amponsah et al50. POAG was the predominant type observed in the present study accounting for 90% of all
cases, this also is consistent with the findings in previous works on glaucoma1,5,12. In fact Yilchung et al13 reported that POAG has the highest prevalence in Africa and that male gender and urban dwellers are more likely to have POAG compared to the female gender and rural dwellers, the reason for this finding is unclear. It has also been documented that males are more likely to have secondary glaucoma following trauma51 , this could not be established in this study because ocular trauma was an exclusion criterion. The mean age of glaucoma subjects is 55.5 ± 11.9 years which is also comparable to the mean age in a previous studies8,52 that documented mean age of 56years52. The highest number of glaucoma subjects was found among the 60-69 years age group, which is in support of earlier documented age specific prevalence
50
of glaucoma1. The reason for this could be that advancing age is a risk factor for developing glaucoma1,53.
The study also set out to see the relationship between the intraocular pressure and the size of the optic nerve sheath. This study shows a statistically significant difference between the mean ONSD of glaucoma group and control subjects. The measured mean ONSD in normal healthy controls is comparable with results obtained from previous studies41 . The study reported a mean ONSD of 4.18mm (SD 0.49) and 4.17mm (SD 0.44) on the right and left respectively. The mean ONSD in glaucoma patients was statistical significantly reduced compared with the mean found in normal healthy controls. This reduction in optic nerve diameter observed in this study in glaucoma is comparable with those of previous
studies15,16,39,43 on ultrasound and MRI. Wang et al43 used MRI to study the eyes of glaucoma patients, they measured the ONSD at 3mm, 9mm and 15mm behind the eye ball. At 3mm behind the eye ball, ONSD of Normal tension glaucoma (NTG) group is significantly narrower than control group (p<0.05). At 9mm, the ONSD of the NTG group and POAG group are both significantly narrower than the control group (p<0.05). At 15mm, the ONSD of the NTG group and POAG group are both significantly narrower than the control group (p<0.05). Contrary to the reduction in ONSD reported in the present study Pinto et al44 reported no difference between the measured ONSD of glaucoma patients and healthy controls. Jaggi et al45 however showed an increase in the ONSD of glaucoma patients. These discrepancies in the ONSD values may reflect differences in patient selection criteria such as age, head position, sample size and most importantly imaging modality. While ultrasound was used in this study, Jaggi used computed tomography and their sample size was only 18 subjects.
This study revealed an inverse, though weak, negative correlation between age and ONSD among glaucoma patients as well as in control, indicating that the reduction in the ONSD could not have been due to the disease process alone, but age may also play a part in the atrophy of
51
the optic nerve. This was also documented in previous study by Beatty et al15. This finding is also consistent with histological studies, reporting an age related decrease in axonal count of the human optic nerve15. This could also be due to older individuals having a more vulnerable optic nerve or have suffered more frequent and prolonged insult to the nerve over their life time, both mechanisms could also play a role in concert to cause atrophy of the optic nerve3. The weak correlation in this study is contrary to what was observed by the aforementioned researcher who recorded a strong correlation, the difference is possibly due to the different geographical area where both studies were carried out and may also depend on the racial factors as the present study was conducted among dark skinned Africans as opposed to the Caucasian population in their study.
The mean LT was not statistically different in the glaucoma and control groups on the right, but a statistically significant difference was observed on the left in this study with a p-value of 0.024. The reason for this unilateral difference cannot be established in this study. From comprehensive literature search, it has been observed that the lens is thickened in closed angle glaucoma11 when compared with controls but this has not been documented in other forms of glaucoma. Furthermore this study has also shown a positive correlation between LT and age among glaucoma patients with a p-value of 0.000 and 0.001 in the right and left eyes respectively. A weak negative correlation was also noted on the left in controls, but no correlation was observed on the right. By the forgoing, it may be possible to predict the lens thickness using the age of patients with glaucoma and vis-versa. Age is however a weak predictor of LT in controls. Hence by linear regression analysis it was demonstrated that age is a significant predictor of lens thickness.
The glaucomatous optic neuropathy has been widely diagnosed; this optic nerve atrophy which is due to loss of retinal fibre layer (RNFL) has also been demonstrated in glaucomatous eyes. It is therefore not surprising to have observed this reduction in the optic nerve sheath diameter in
52
this study among glaucoma patients compared with healthy normal subjects. The result of this study should assist the clinician in further evaluation of glaucoma patients and also assist in screening of glaucoma suspects, especially in the setting of an opaque media where examination of the fundus might prove difficult or impossible.
LIMITATIONS
Ultrasound being user dependent, may give rise to minor but negligible variation in values obtained. Moreover all measurements were done by the same researcher; hence inter-observer
53
reliability was not evaluated. Most researchers used ultrasound machines with probe frequency of 20MHz, this study however utilized a 6.5-12MHz machine, and the effect of this was expected to be negligible as frequencies beyond 10MHz have shown equally reliable results.
54
CONCLUSION
Real time B-mode ultrasound scan has been shown by this study to be a useful tool in determining the ONSD as well as LT in normal and glaucomatous subjects. It also demonstrated the reduction of the ONSD in glaucomatous eyes. Hence it can be an alternative tool in the assessment of eyes of patients with glaucoma as well as screening for glaucoma suspects where the traditional triad of tonometry, visual field testing and optic nerve evaluation are impossible or not available.
55
RECOMMENDATION
In resource poor regions like ours, B-mode ultrasound could be recommended as the initial screening method for glaucoma suspects. Further studies to ascertain possible reason for the differences in our result of ONSD values against those reported in previous literatures is also recommended.
56