The prevalence of diabetic retinopathy in this study was 42.2% which is higher than most of the results of previous studies done in Ethiopia and other African countries. 5,6,8,14-18 The high prevalence seen in our study could be due to the fact that the sample population was taken from a retina subspecialty clinic where most of the patients were referred from the medical diabetic clinic for visual complaints unlike the studies mentioned above which were done at medical diabetic clinics. Different sampling techniques, sample size and diagnostic method may have also contributed to this difference. The very low prevalence (13%) seen in the Arbaminch General Hospital study in Ethiopia by Chisha Y et.al may be due to the fact that a majority of patients had a diabetes duration below 6 years and age less than 60 years as well as the retrospective record review design of the study. 17
The prevalence in the current study was also higher than the findings from New Zealand, Spain and USA where rates of 22.5%, 14.9% and 14.7% were reported respectively. 18,19,20 This difference could be due to the very large sample size in those studies, the difference in economic status between Ethiopia and the countries in which those studies were conducted affecting patient care. 8
However, the finding in our study is lower than the prevalence reported by Shibiru T et.al (51.3%) from Tikur Anbessa Hospital in Ethiopia. 7 The variations in sample size and diagnostic method may have caused this discrepancy. Some studies from other parts of the world also reported figures higher than the prevalence seen in the current study. 21-24 Many factors may have contributed to this difference including variations in sampling techniques, sample size, study setting, methods of screening, level of awareness among study participants, level of glycemic control and diabetic patient care.
A majority of the study population was constituted by type 2 DM patients (88.4%) in this study, a trend similar to studies done at Tikur Anbessa hospital Ethiopia, 5 Jimma University hospital Ethiopia 6 and Arbaminch General hospital Ethiopia10 where type 2 DM constituted 53.6%, 72.8% and 74.1% of the study subjects respectively.
The prevalence of diabetic retinopathy in type 1 DM patients (46.2%) was slightly higher than the prevalence among type 2 DM patients (41.7%) but the difference was not statistically significant. This is in line with the findings reported from Cameroon, Zimbabwe and Kenya where no significant difference was noted in the prevalence of diabetic retinopathy between the two groups. 15,16,21
Although there was no statistically significant difference between males and females in the prevalence of diabetic retinopathy, the prevalence was higher in males (46.7%) compared to females (35.6%). Similar finding was seen in the study done in Cameroon by Njikam E.J et.al in 2011 where the prevalence of diabetic retinopathy in males and females were found to be 54.2% and 46.3% respectively, with no statistically significant association between diabetic retinopathy and gender.21 This was in contrast to the report from Kenya by Mariangela W.N in 2011 where prevalence of 19.7% and 37.4% were reported in males and females respectively, and showed significant association of diabetic retinopathy with female gender.16 This difference might be partly due to the large number of female patients in the later study.
Longer duration of diabetes was significantly associated with the occurrence of diabetic retinopathy in this study and patients with disease duration of 6 years or more were more likely to develop diabetic retinopathy (AOR= 2.91: 95% CI; 1.01 - 8.35) as compared to those with disease duration of less than 6 years. This finding was consistent with major global meta-analyses and most of the studies done in other African countries. 14-16,21,25-27
The mean age of diabetes patients in this study was 55.4(±13.5) years which was higher than studies done in Ethiopia 5,6 and similar to studies done in other parts of the world. 14-16,21,26,27 However, the mean age of patients with DR (53.23 ± 13 years) was lower than those without DR (57.11 ± 13.67 years) in our study. Patients who were <60 years of age constituted 56% of the participants and they were more affected by diabetic retinopathy than those ≥ 60 years of age with statistically significant difference (AOR= 3.2: 95%CI; 1.19 - 8.63) contrary to many of the other studies that showed advanced age was associated with diabetic retinopathy. 5,6,17,18,19,22
There was correlation of diabetic retinopathy with the form of therapy in this study, and diabetic patients who were on insulin alone or combined Insulin and OHA therapy had higher prevalence of diabetic retinopathy. This was in contrast to the report in Cameroon (2011) which showed higher prevalence in those on OHA therapy.5 With increasing duration of diabetes and advancing age, patients with type 2 DM may be unable to control their blood sugar level with OHA only and these patients often start insulin alone or combination of insulin and OHA therapy to improve glycemic control. This Poor glycemic control may have contributed for the high prevalence of diabetic retinopathy in the group of patients who were on insulin alone or combination of insulin and OHA therapy in our study.
Poor glycemic control is a risk factor for the development and progression of diabetic retinopathy and is associated with higher prevalence of diabetic retinopathy as shown by reports from different studies. Due to unavailability of HbA1c test, which is the best indicator of the level of glycemic control in the few months preceding the test, fasting blood Glucose (FBG) level at the time of data collection was used to assess the level of glycemic control in our study.
Only 35.1% of study patients had their FBG level below 126 mg/dl and the overall mean FBG level was 157.68 ± 64.5 mg/dl. Diabetic patients with diabetic retinopathy had slightly higher mean FBG level (160.86 ± 70.6 mg/dl) than those who had no diabetic retinopathy (155.35 ± 59.9 mg/dl) but there was no significant association between FBG level and diabetic retinopathy. This was contrary to the finding reported by Sharew G et.al in Jimma university hospital Ethiopia (2009) which showed that FBG level was significantly associated with diabetic retinopathy.9 This may be due to the fact that patients with diabetic retinopathy in the Jimma university hospital study had poor glycemic control with relatively much higher FBG level than those without diabetic retinopathy at presentation compared to the findings in our study. Association of poor glycemic control with diabetic retinopathy was also shown by studies done in other parts of the world.21,28-31 This difference in the findings between our study and other studies may have resulted from the use of HbA1c to assess level of glycemic control in the studies mentioned.
The mean systolic blood pressure of patients with diabetic retinopathy in this study was 130.56 ± 15.7mmHg which is slightly higher than that of patients without diabetic retinopathy (127.73 ± 13.4 mmHg). It is well established that systemic hypertension affects development and progression of diabetic retinopathy in patients with diabetes. Our study showed significant relationship between systolic blood pressure ≥ 140 mmHg and the occurrence of diabetic retinopathy. This correlates well with research findings from other settings in Ethiopia 5,6,7,17 as well as elsewhere. 21,26,27
Majority of patients with retinopathy had NPDR (38.6%) which was similar to figures reported from Tikur Anbessa Hospital, Ethiopia (36.1%) and Jimma University Hospital, Ethiopia (38.9%) but higher than those reported from Kenya (25.7%) by Dr. Mariangela W.N and Nigeria (24%) by Lawan A et.al. Diabetic macular edema was seen in 13(5.7%) patients of whom 6 (2.7%) had CSME which is lower than Jimma University Hospital study in Ethiopia, (5.5%), the 2011 Kenyatta National Hospital study (4.2%) and 2011 Yaoundé Central Hospital study (8.1%). The prevalence of PDR in this study was 3.6% which is higher than those reported from studies done in Ethiopia 5,6 but lower than those reported from other African countries, 5.9% in Kenya (2011), 3.7% in Nigeria (2009) and 14.3% in Cameroon (2011).
Vision threatening diabetic retinopathy was seen in 24 (10.7%) patients which is comparable with results from Zimbabwe (11.4%), Kenya (11.9%), and Uganda (14.6%) among studies done in Africa. 15,16,32 Slightly higher findings were reported by Sultan S et.al 33 and Al-Rubeaan K et.al 34 which showed VTDR in 17.6% and 16.3% respectively. The prevalence of VTDR in our study was also lower than that was seen in Cameroon (27.3%) which may be due to the high prevalence of diabetic retinopathy and associated poor glycemic control in the Cameroon study. 5 Some studies reported lower prevalence of VTDR than ours. 6,35-39 This discrepancy could be a result of variations in sample size, sampling techniques and studies setting as most of these studies were population based.
One of the limitations of our study is the relatively modest sample size but similar studies done in African health care settings rarely have larger sample size. Another limitation is our inability to determine HbA1C level due to unavailability of the test at the study center. As a result we took only a single measurement of Fasting Blood Glucose level to assess the level of glycemic control in our patients and that is not the idea way of measuring the glycemic status of patients.