Visual Impairment and Its Associated Factors among People Living with Type-2 Diabetes Mellitus at Dessie Town Hospitals, Northeast Ethiopia, 2020

Abstract

Visual impairment (VI) is a functional limitation of the eye(s) that results in reduced visual acuity, visual field loss, visual distortion, perceptual difficulties, or any combination of the above. Type-2 diabetes mellitus (T2DM) is one of the common causes of VI. The current study aimed to determine the prevalence and predictors of VI in diabetes individuals. Institution-based cross-sectional study was carried out, and VI was measured using visual acuity test. We used Epi Data 3.1 and SPSS - 21for data entry and statistical analysis, respectively. To find statistically linked factors of VI, we used both binary and multivariable logistic regression. The strength of association was estimated using AOR at 95% CI. Statistical significance was declared at p less than 0.05. The present study revealed 37.58% of people have VI, which is statistically linked to age, lack of regular exercise, diabetes for >5 years, insulin treatment, and poor glycemic control. Finally, individuals with T2DM who had VI accounted for more than a third of those treated in Dessie town hospitals. Advanced age, poor frequent exercise, longer duration of diabetes, and insulin are predictors. To lower the risk of VI and visual loss, early identification of VI through screening and regular follow-up is recommended.

1. Introduction

1.1 Statement of the problem

Type-2 diabetes mellitus (T2DM) is a global metabolic disease, mainly seen in developing countries [1]. Patients with T2DM are highly prone to have cardiovascular disease, renal failure, neurological problems, and retinopathy [2, 3]. If the vision of an individual has been lost once due to diabetic retinopathy, usually it could not be restored, despite some forms that probably may be treated by complex vitreoretinal surgery [1, 2, 3].

Visually impaired people (VI) could not maintain employment, productivity, independence, their medical expense, or they may not be able to legally licensed to drive, they get difficulties in reading and fail to work inorder to obtain wages to their family and execute their social responsibilities as a whole [4, 5]. It also has been linked to falls, injury, and worsened status in spanning mental health, cognition, social function, and educational attainment [6]. People developing visual impairment(s) encounter a significant challenge. They experience major life changes, such as general health limitations or loss of nearby family member [1, 6]. They also are at higher risk of violence and abuse, which limit them from participating in and contributing to their families and societies on an equal basis with others [7].

Globally, prevalence of visual impairment increased sharply from 441.1 million [8] to 2.2 billion [4]. Of which at least 1 billion of VI could have been prevented. Among the causes of visual impairment, 1% was associated with diabetic retinopathy [9]. Diabetic retinopathy is damage to blood vessels in the retina, swelling in the central part of the retina, and abnormal blood vessels formation, which can lead to bleed or cause scarring of the retina that results in visual impairment [10].

Sub-Saharan Africa report on visual impairment figures out that the prevalence of visual impairment was 29.7% among diabetes mellitus [11]. Cataract, diabetic retinopathy, glaucoma, and maculopathy were the causes of visual impairment. Ethiopian systematic review and meta-analysis report revealed that the prevalence of diabetic retinopathy among diabetes mellitus patients was 19.48%, which is one cause of preventable visual impairment among middle-age and elderly populations [12].

In Ethiopia, no research is done on visual impairment among diabetes mellitus patients. But one previous study at the same setting found that visual disturbance was the second major chronic complication among newly diagnosed diabetes mellitus patients that was detected by clinical findings and questionnaire-based approaches. Moreover, specific associated factors for developing visual impairment had not been assessed. In parallel to rapid increment of DM, visual impairment is still public health challenge due to its chronic impacts.

Therefore, this study is designed to assess the magnitude and associated factors of long-term complication of DM, particularly vision impairments among both newly diagnosed and old type 2 diabetes mellitus patients having follow-up in Dessie town hospitals, northeastern Ethiopia, using ocular examinations and ocular tests. Since diabetes mellitus is a lifelong disease, the findings of this study will be used as input for all stakeholders such as governments, healthcare providers, people living with DM, civil society, food producers, and manufacturers and suppliers of medicines and technology. Collectively, they can make a significant contribution to halt the most serious and feared diabetes mellitus complication, (i.e., vision loss) and improve their lives as well.

1.2 Literature review

1.2.1 Definition and classification of visual impairment

Visual impairment (VI) is a functional limitation of the eye(s) due to a disorder or disease that results in reduced visual acuity, visual field loss, photophobia, diplopia, visual distortion, visual perceptual difficulties, or any combination of the above [13]. Visual impairment is also often defined as presented visual acuity of worse than either 20/40 0r 20/60 to no light perception (NLP) in either or both eyes [14].

Visual impairment can be congenital or hereditary. Other main causes of VI incude refractive error, cataract, glaucoma, corneal opacity, age-related macular degeneration, and diabetic retinopathy [4, 13, 15, 16, 17]. It is also linked to ocular infection or disease, trauma, and systemic diseases such as hyperthyroidism [18], rheumatoid arthritis, HIV/AIDS, and hypertension [19].

Visual impairment in category Inernational Classification of Disease code H54 (H54) comprises from category 0 to 9 based on presenting distance on visual acuity test and classified as low vision (mild, moderate, and severe visual impairments) and blindness [4, 20] (annex VI).

1.2.2 Pathogenesis of visual impairment due to diabetes mellitus

Chronic hyperglycemia following long-standing DM is claimed to cause visual impairment via the production of inflammatory factors, which lead to inflammation of endothelium that in turn reduces the integrity of the blood retinal barrier in diabetic eyes [21, 22]. These are attributed to decreased in the activity of nitric oxide, increased in the activity of angiotensin II, endothelin-1, and vascular endothelial growth factor (VEGF) [3]. Besides, there is slow growth of new blood vessels in the iris and trabecular meshwork, which inhibits outflow of the aqueous humor fluid that causes irreversible damage to the optic nerve, eventually leading to blindness [21]. Disruption of the blood retinal barrier is responsible for developing retinovascular diseases including diabetic retinopathy (DR). Henceforth, DR causes vascular leakage and macular edema. If timely management is not tailored, there will be reduced vision and eventually blindness [2, 23].

Diabetes-associated glucose toxicity leads to biochemical changes due to polyol pathway and activation of protein kinase C (PKC) [24]. The polyol pathway is started by the conversion of glucose to sorbitol by the aldose reductase, and then sorbitol is changed into fructose by sorbitol dehydrogenase [25]. Accumulation of sorbitol leads to osmotic changes resulting in hydropic lens fibers that degenerate lens and form sugar cataracts [26]. Both polyol pathway and PKC result in increased oxidative stress, inflammation, and vascular incompetence. Oxidative stress and inflammation cause upregulation of growth factors that play a role in the breakdown of the blood retinal barrier and development of macular edema [26, 27]. Chronic hyperglycemia also increases diacylglycerol (DAG) that leads to the activation of protein PKC. Then PKC increases vascular permeability and upregulation of retinal vascular endothelial growth factors. Diacylglycerol and PKC pathways progressively affect inflammation, neovascularization, and retinal blood flow, which ends up with DR and progressively visual impairment [27, 28].

1.2.3 Prevalence of visual impairment among type-2 diabetic mellitus patients

Pandemicity of T2DM is rising rapidly probably due to increasing obesity, reduced physical exercise as countries become more industrialized and aging of the population [6]. Between 2010 and 2030, a 20% increase DM cases in developed countries and a 69% increase in developing countries have been predicted [29]. According to WHO report, currently global prevalence of visually impaired people is estimated to be 2.2 billion. Of these, at least 1 billion visual impairment is preventable or has yet to be addressed [4].

Many studies were conducted in different countries on the prevalence of visual impairment among T2DM. VI among T2DM in Sankara Nethralaya was 4% [30], Peru 26.3% [31], China 10% [16], Jordan 17.7% [32], Turkey 16.2% [33], and Yemenian 76.5% [34].

Studies in Africa reported different magnitude of VI. The prevalence of unilateral visual impairment among DM patients was 78.25% in South Africa [35], 17.1% in Zambia [36], and 18.4% in Kumasi, Ghana [37]. In Nigerian, 24.1% of T2DM cases had visual impairment [38]. Moreover observational studies in Tunisia [39] and Cameron [40] revealed 22.2% and 22.6% prevalence of VI among DM patients, respectively.

In Ethiopia, there is no visual impairment study done among diabetes mellitus patients, but in a study at St. Paul’s Hospital, visual impairment was 17.6% among all ophthalmic cases. Of this, 58.7% had low vision and 41.3% had blindness [23] and that of Dessie referral Hospital, the prevalence of visual disturbance among diabetes mellitus patients was 28.9% [41].

1.2.4 Associated factors of visual impairment among T2DM patients

Factors associated for developing visual impairment vary among different studies. A study done in China, Jordan, and Yemen revealed that visual impairment was statistically associated to patients’ age, duration of diabetes, body mass index (23.4 ± 3.9), education level (≤ primary education), and insulin as treatment option [16, 32, 42].

Finding from Sankara Nethralaya showed that visual impairment was higher among diabetes patients age > 60 years and low socioeconomic status [30]. In Peruvian population, hypertension, hemoglobin A1c, and use of insulin or sulfonylurase as diabetic treatment were associated to visual impairment [31]. Diabetes and visual impairment study in sub-Saharan Africa (evidence from Cameroon) revealed that factors such as age (≥ 50), duration (≥10 years), and hypertension were associated with severe visual impairment [11]. A study from Zambia, Turkey, and Tunisia findings indicated that age, overweight, duration of diabetes, high random blood sugar, high systolic blood pressure, and insulin as diabetic treatment were significantly associated to visual impairment [33, 36, 39].

In Arbaminch Referral Hospital, baseline age (≥ 60), duration of diabetes (≥6), baseline systolic BP level (>140) were significantly associated to DR [43] while in Dessie, patients’ age, drug regimen, and specific medications taken were associated with diabetic complication (Figure 1) [41].

Figure 1.

Conceptual framework adapted from different literatures illustrating possible factors affecting visual impairment among people living with T2DM at Dessie, Northeast Ethiopia, 2020 [11, 16, 31, 32, 36, 39, 43].

1.2.5 Developed conceptual framework

1.3 Justification of the study

To the best of my knowledge, there was no study conducted in Ethiopia to determine the prevalence of VI and its associated factors among people living with T2DM. Vision loss is the most serious and feared outcome of DM-associated complication. After completion, this study will provide an in-depth and comprehensive information on the prevalence of VI and identify associated factors among people living with T2DM. The findings of this study will enable DM patients to implement preventive strategies and adhere to self-care management and medication to improve glycemic control and hence preventing VI by targeting associated factors. Preventive strategies include screening for DR, provision of follow-up care, and ensuring the follow-up techniques that meet the standard clinical guidelines. Making available an adequate referral mechanism allows all patients to screen and diagnose DM early so that possible ensuring eye complications are detected. Indeed, it will create an effort to develop effective health service programs and policies for better management of DM and cost-effective strategies in Ethiopian context as well as a baseline for the coming researchers and stakeholders at higher level.

2. Objectives

2.1 General objective

• The aim of the current study was to assess the prevalence of visual impairment and its associated factors among people living with T2DM at Dessie Town Hospitals, Northeast Ethiopia, 2020

2.2 Specific objectives

• To determine the prevalence of VI among people living with T2DM

• To identify associated factors of VI among people living with T2DM

3. Methods and materials

3.1 Study setting and period

This study was conducted at Dessie town hospitals (both government and private hospitals). Dessie town is located 400 km away from Addis Ababa, the capital city of Ethiopia. During the study period, there were four private and one governmental referral hospitals in Dessie town. Current report of Dessie zonal administrative office showed that these hospitals engaged in providing service for more than 3.5 million people. According to annual summative report of all diabetic clinics in Dessie town hospitals, an estimated number of 14,000 people living with T2DM were served. In each hospital, there was diabetic follow-up clinic for treatment and follow-up of people living with DM. The follow-up date in each hospital was from Monday to Friday. The actual data collection period was from February 15, to March. 152,020.

3.2 Study design

Institution-based cross-sectional study design was employed.

3.3 Source population

Source population was all people living with type-2 diabetes mellitus (both old and newly diagnosed cases) came to diabetic clinic of each hospital.

3.4 Study population

Study population was all people living with type-2 diabetes mellitus visiting diabetic clinic of each hospital during data collection period.

3.5 Eligibility criteria

3.5.1 Inclusion criteria

People living with T2DM who had follow-up and newly diagnosed T2DM patients in diabetic clinic during study period were included in the study.

3.5.2 Exclusion criteria

People living with T2DM who were seriously ill, pregnant women, and those who had HIV/AIDS, trachoma, acute eye infections (bacterial or viral), trauma to the eye, history of head injury, and history of stroke were excluded from the study.

3.6 Sample size determination and sampling technique

Sample size was determined using a single population proportion formula by considering the following assumptions:

Prevalence of visual impairment =28.9% [41], 95% confidence interval (zα/2). and 5% margin of error (d = 0.05). Then sample size (n):

$$\mathrm{n}=\frac{{\left(\mathrm{Z\alpha }/2\right)}^2\mathrm{p}\left(1-\mathrm{p}\right)}{{\mathrm{d}}^2}.$$E1

$$\mathrm{n}={\left(1.96\right)}^2\times \left(0.289\right)\times \left(0.711\right)/{\left(0.05\right)}^2=316$$E2

Adding non-response rate of 5% yielding a final sample size of 332.

Sample size allocation.

Sample size for each hospitals was allocated by using sample size allocation formula.

$${\mathrm{n}}_{\mathrm{j}}=\frac{\mathrm{n}}{\mathrm{N}}{\mathrm{N}}_{\mathrm{j}}$$E3

Where n_j = sample size for each hospitals.

n = total sample size.

Nj = total people living with type 2 diabetes in each hospitals.

N = total people living with type 2 diabetes in all hospitals.

$${\mathbf{n}}_{\mathrm{DRH}}=\frac{332}{14000}\ast 10210=242$$E4

$${\mathbf{n}}_{\mathrm{SRH}}=\frac{332}{14000}\ast 1440=34$$E5

$${\mathbf{n}}_{\mathrm{ERH}}=\frac{332}{14000}\ast 1250=30$$E6

$${\mathbf{n}}_{\mathrm{BRH}}=\frac{332}{14000}\ast 1100=26$$E7

Where DRH = Dessie Referral Hospital, SRH=Selam General Hospital, ESH = Ethio General Hospital and BGH=Bati General Hospital (Figure 2).

Figure 2.

Diagrammatic representation of sampling procedure for T2DM at Dessie town Hospitals, Northeast Ethiopia, 2020.

In this study, Dessie Referral Hospital and all private hospitals were included and simple random sampling technique using lottery methods was applied to recruit study participants from those of type 2 DM cases who came to the clinic for follow-up and newly diagnosed T2DM patients during study period.

3.7 Study variables

3.7.1 Dependent variable

• Visual impairment

3.7.2 Independent variables

• Sociodemographic variables: Age in years, sex, monthly income, marital status, and educational level

• Behavioral variables: Diet, exercise, and diabetic regular follow-up to T2DM

• Clinical-related variables: Glycemic control, plasma glucose level, duration of DM, comorbidities (hypertension and obesity), and treatment option for people living with type 2 diabetes mellitus.

3.8 Operational definitions

Visual impairment: Is any loss or abnormality in an anatomical structure or a physiological or psychological function [13]. It is presenting visual acuity of worse than either 20/40 0r 20/60 to no light perception (NLP) in either or both eyes, which includes both low vision and blindness [44].

Low vision: Even with corrective lenses, it is inability to clearly see at a distance of 6 meters (20 feet) that individuals with normal vision can clearly see at a distance of 12 meters (40 feet) or visual acuity between 6/12 and 3/60. Low vision includes mild visual impairment (VA between 6/12 and 6/18), “moderate visual impairment (VA between 6/18 and 6/60)” and “severe visual impairment (VA between 6/60 and 3/60)” from all causes [4].

Blindness: Inability to read the largest letter on a vision chart at a distance of 3 meters (10 feet) or visual acuity was worse than 3/60 [45].

Visual acuity/VA: Simple, noninvasive measure of the visual system’s ability to discriminate two high-contrast points in space. It is usually taken at a distance of 6 m or 20 feet [9]. It is expressed in terms of A/B, where A: distance b/n observer and letters and B: expected distance that the health eye can observe.

Obesity: A BMI of ≥30 kg/m² is considered obese. Class 1 obesity (BMI: 30–34.9 kg/m²) is one of the most common subtypes. Obesity class 2 (BMI 35–39.9 kg/m²), extreme obesity or morbid obesity (BMI 40 kg/m²) is classified as class 3 (BMI 40 kg/m²) [46].

Exercise: Exercising for less than 150 minutes per 3–5 days/week is regarded poor, while exercising for more than 150 minutes per 3–5 days per week is considered good [47].

3.9 Data collection procedures and tools

We used semistructured interviewer-administered questionnaire, tape meter, and weight balance to collect data. Type of diabetes already diagnosed by physician and their ID list were used to select study participants. After obtaining consent, weight, height, and blood pressure were taken. Then they were scheduled for eye examination at the eye clinic of the same institution and had additional benefits of individualized counseling, care, and referral depending on the ocular findings. At the eye clinic, each patient had visual acuity assessment with illuminated Snellen’s chart for each eye at 6 m. Training was given by principal investigator about the objective of study, data collection techniques, and ethical issues had been given to four data collectors who were BSc ophthalmic nurses and clinical nurses and one supervisor prior to actual data collection. Pre-test was applied on 17 patients in Borumeda General Hospital for evaluation of consistency, approachability, and feasibility of the questionnaire. Information obtained was strictly kept confidential.

3.9.1 Ophthalmic examination

Ophthalmic examination was performed by ophthalmic nurse and ophthalmologist. Careful ocular history, inspection, and examination of the eye using slit-lamp and dilated fundus examination were conducted. Visual acuity (using snellen chart), macular degeneration (ophthalmoscope), intraocular pressure (using tonometry), and history of night blindness for vitamin A deficiency were taken.

3.9.2 Snellen chart for visual acuity test

Visual acuity is measured by taking 6 m notation. Visual acuity was performed in a properly illuminated quiet room, using Snellen chart at 6 m to discriminate different letters. Each eye was tested separately, and the procedure was repeated, then the best average was taken. The person who could identify the letters of the size 6 at 6 m (20 at 20 feet) was said to have normal vision. The numerator expresses the distance between the observer and the letters while the denominator expresses the distance at which the letter could be distinguished by the normal eye.

3.10 Data quality control

Data quality assurance was maintained starting from design. The questionnaire was first prepared in English and then translated from English to Amharic (local language) and retranslated to Amharic by another expert to ensure understanding of the items for the participants and its consistency. Pre-test was done on 17 people living with T2DM at Borumeda Hospital. Training was given for data collectors and supervisor on the data collection tool and ethical issues during data collection.

3.11 Data processing and statistical analysis

After completing the data collection process, data were entered into Epi data-3.1 by data entry clerk, then exported into SPSS version 22 for analysis. Data completeness, consistency, and outliers were checked. Continues data were described by median and inter-quartile range while frequency with percent was used to describe the results of categorical variables. Then results were presented using tabulation, graph, and charts. Uni-variable analysis was used to describe independent variables, and bi-variable binary logistic regression analysis was performed to select potential candidate variables for the final model with cutoff point of p value ≤0.25 [48]. Model fitness was checked by Hosmer and Lemeshow goodness of fit test. Multivariable binary logistic regression analysis was done to identify significant factors of visual impairment. Adjusted odds ratio with 95% CI was computed to show significant factors. In the final model, variables with a p value ≤0.05 were considered as statistically significant.

4. Ethical consideration

The institutional review board (IRB) of the University of Gondar, College of Medicine and Health Sciences, granted ethical approval with reference number 1839/02/2020. Prior to data collection, Dessie Town Hospitals provided an official authorization letter. After describing the goal of the study to each participant, they signed a written informed consent form. Participants had full mandate to participate or to refuse even to withdraw at any time they want from the study. The information obtained was kept confidential.

5. Dissemination plan of results

Electronic copy of the thesis will be published by University of Gondar for online access. We also plan to disseminate the result of this study as a copy of the document to Dessie town private hospitals and Dessie Refferal Hospital. Attempts will be made to present the findings at various scientific conferences, workshops, and meetings. In addition, an effort will be made to publish the findings in a peer-reviewed scientific journal.

6. Results

6.1 Sociodemographic characteristics of participants

Out of the total of 332 study participants, 322 have participated in the study yielding a response rate of 97%. The median age of participants was 52 years (IQR: 45–60 years) ranging from 24 to 87 years. One-hundred seventy and five (54.3%) study participants were male giving female-to-male ratio of 1: 1.18. Two hundred and twelve (65.8%) individuals were Islamic religion followers. Seventy-three (22.7%) participants were unable to read and write. One-hundred and fifteen people (35.7%) worked for a private company, 291 (90.4%) were married, and 249 (77.3%) lived in a town. The median monthly income of the participants’ households was 3570 ETB (IQR: 2000–5195, Min = 800, Max = 9600ETB) (Table 1).

Variables	Categories	Frequency	Percent
Age in years	20–40	37	11.5
	41–59	189	58.7
	60–87	96	29.8
Sex	Male	175	54.3
	Female	147	45.7
Religion	Muslim	212	65.8
	Orthodox	103	32
	Others*	7	2.2
Marital status	Never married	31	9.6
	Married	291	90.4
Residence	Urban	249	77.3
	Rural	73	22.7
Educational level	Unable to read and write	73	22.7
	Primary	84	26.1
	Secondary	79	24.5
	Diploma and above	86	26.7
Occupation	Government workers	70	21.8
	Private workers **	115	35.7
	Farmer	37	11.5
	House wife	60	18.6
	Others***	40	12.4

Table 1.

Sociodemographic characteristics of people living with T2DM at Dessie town Hospitals, Northeast Ethiopia, 2020 (n = 322).

Other*: protestant and catholic, Private worker**: construction, daily laborer, driver, mechanic, merchant, Others***: jobless, pensioner.

6.2 Type-2 DM and vision-related characteristics of participants

Forty-six (14.3%) people living with T2DM had trouble in adjusting light while entering from bright to dim light. One-hundred and seventy three (53.7%) of DM patients had duration 5 years and less since diagnosis while 24 (7.5%) were new cases. Of all participants, two-hundred and sixty-four (82.0%) T2DM patients had regular follow-up in clinics. One-hundred and ninety nine (61.8%) of participants had poor or inadequate physical exercise. In 201 (62.4%) T2DM patients, the treatment option for DM was oral hypoglycemic agent without insulin. Thirty-seven (11.5%) participants were obese (BMI: median = 34.8, IQR: (23.1–27.65) and 98 (30.4%) individuals had co-morbid hypertension. Respondents had a median with IQR of baseline random plasma glucose (311, (260–396)) and fasting plasma glucose (160, (140–208)), respectively. One-hundred and eighty-seven (58.1%) participants had poor glycemic control (Table 2).

Variables	Categories	Frequency	Percent
Have trouble in adjusting light (night blindness)	Yes	46	14.3
	No	276	85.7
Regular follow-up	Yes	264	82
	No	58	18
Duration of diabetes (in years)	Newly diagnosed	24	7.5
	≤5 years	173	53.7
	>5 up to 24 years	125	25.8
Treatment (n = 296)	OHA without insulin	201	67.9
	Both of OHA and insulin	73	24.7
	Insulin only	22	7.4
Regular exercise	Good*	123	38.2
	Poor*	199	61.8
BMI (kg/m^2)	Underweight	5	1.5
	Normal	159	49.4
	Overweight	121	37.6
	Obese	37	11.5
Comorbid hypertension	Yes	98	30.4
	No	224	69.6
Glycemic control	<152 mg/dl (good)	135	41.9
	≥152 mg/dl (poor)	187	58.1

Table 2.

Factors related to diabetes mellitus and vision-related characteristics of participants at Dessie town Hospitals, Northeast Ethiopia, 2020 (n = 322).

OHA: oral hypoglycemic agents, Good*: doing exercise every other day ≥30 min/day, Poor*: do not doing exercise at all or doing exercise <30 min/day.

6.3 Prevalence of visual impairment among people living with T2DM

In the current study, the prevalence of visual impairment was 37.58% [95% CI: 32.3–42.9] with mean ± SEM (0.38 ± 0.027) (Figure 3).

Figure 3.

Pie chart showing prevalence of visual impairment among people living with T2DM at Dessie town Hospitals, Northeast Ethiopia, 2020 (n = 322).

Fifty-eight (18.0%) participants with visual impairments were in the moderate visual impairment category. Among the overall prevalence of visual impairment, 43(13.4%) had bilateral vision impairment and 78(24.2%) had monocular vision impairment out of total participants. Of all visually impaired T2DM, 107 (33.2%) and 14 (4.4%) had low vision and blindness, respectively (Table 3) (Figure 3).

Visual impairment category		Frequency	Percent
<6/12–6/18	Bilateral mild VI	28	23.1
<6/18–6/60	Bilateral moderate VI	10	8.3
<6/60–3/60	Bilateral severe VI	1	0.8
<3/60-NLP	Bilateral blindness	4	3.3
<6/12–6/18, other eye 6/6–6/12	Monocular mild VI	0	0.0
<6/18–6/60, other eye 6/6–6/18	Monocular moderate VI	48	39.7
<6/60–3/60, other eye 6/6–6/60	Monocular severe VI	20	16.5
<3/60-NLP, other eye 6/6–3/60	Monocular blindness	10	8.3
Total		121	100

Table 3.

Forms of visual impairment categories among people living with T2DM at Dessie town Hospitals, Northeast Ethiopia, 2020, (n = 322).

Note: NLP-no light perception, VI-visual impairment.

Visual impairment among people living with T2DM increases with age. Among over all T2DM patients participated in this study, three (0.93%) participants aged 20–39 years, 62 (19.25%), individuals aged 40–59 years, and 56 (17.39%) participants aged 60–87 years were visually impaired (Figure 4).

Figure 4.

Distribution of visual impairment among different age groups of people living with T2DM at Dessie town Hospitals, Northeast Ethiopia, 2020 (n = 322).

In T2DM patients, visual impairment increases with duration of diabetes. Among all T2DM participants, 1 (0.31%) new T2DM cases, 42 (13.04%) participants with 5 years and below duration, and 78 (24.22%) individuals with >5 up to 24 years of duration were visually impaired (Figure 5).

Figure 5.

Magnitude of VI across duration of people living with T2DM at Dessie town Hospitals, Northeast Ethiopia, 2020 (n = 322).

6.4 Associated factors of VI among patients living with T2DM

Age, sex, marital status, educational level, occupation, residence, regular exercise, duration of DM, treatment option, baseline random plasma glucose, hypertension, and glycemic control were candidates for final model. In multivariable analysis age, poor regular exercise, duration of diabetes, insulin treatment, and poor glycemic control were statistically significant with visual impairment.

The odds of having visual impairment for each age increase of a unit (a year) were 1.06 times (AOR: 1.06, 95% CI: 1.02, 1.09). Participants who relied on insulin were 14 times (AOR = 14.05, 95% CI: 2.72, 72.35) more likely to get visual impairment than those who used treatment options without insulin. The odds of having visual impairment in diabetes mellitus who had poor physical exercise were 2.91 times (AOR = 2.91, 95% CI: 1.47, 5.76) more likely than those who were good in physical exercise. People living with T2DM with duration of more than 5 years were 2.42 times (AOR: 2.42, 95% CI: 1.24, 4.73) more likely to acquire visual impairment than those with duration of 5 years and lower. Those who had poor glycemic control were 2.17 times (AOR: 2.17, 95% CI: 1.13, 4.14) more likely to develop visual impairment in contrary to good glycemic control (Table 4).

Variables	Categories	VI		COR (95% CI)	AOR(95% CI)
		No	Yes
Sex	Male	119	56	1	1
	Female	82	65	1.68 (1.06, 2.65)	1.29(0.56, 2.96)
Marital status	Unmarried	27	4	1
	Married	174	117	4.54(1.54, 13.31)	1.25(0.30, 5.11)
Education	Unable to read and write	32	41	1	1
	Primary (1–8)	52	32	0.48 (0.25, 0.90)	0.92 (0.33 2.50)
	Secondary	51	28	0.43 (0.22, 0.82)	1.11 (0.37, 3.29)
	Diploma and above	66	20	0.23 (0.12, 0.46)	0.67 (0.20, 2.23)
Occupation	Government worker	52	18	1	1
	Private worker	81	34	1.21 (0.62, 2.36)	0.80 (0.34, 1.88)
	Farmer	23	14	1.75 (0.74, 4.12)	0.35(0.10, 1.17)
	House wife	26	34	3.77 (1.80, 7.92)	1.54 (0.59, 4.00)
	Other	19	21	3.19 (1.40, 7.25)	1.33(0.42, 4.17)
Residence	Urban	166	83	1
	Rural	35	38	2.17(1.27, 3.68)	1.85(0.71, 4.80)
Regular-	Good	94	29	1
exercise	Poor	107	92	2.78(1.68, 4.59)	2.91(1.47, 5.76)**
Duration of diabetes	≤5 years	154	43	1
	>5–24 years	47	78	28.59(3.68, 221.75)	2.42(1.24, 4.73)**
Treatment	OHA without insulin	140	61	1
	Both of OHA and insulin	34	39	2.63(1.52, 4.56)	1.45(0.71, 2.97)
	Insulin only	2	20	22.95(5.20, 101.25)	14.05(2.72, 72.35)**
Hypertension	Yes	51	47	1.86(1.15, 3.03)	1.26(0.65, 2.44)
	No	150	74	1
Glycemic-	<152 (good)	94	41	1
control(FPG)	≥152 (poor)	107	80	1.71(1.07, 2.73)	2.17(1.13, 4.14)*

Table 4.

Multivariable binary logistic regression analysis for associated factors of visual impairment among people living with T2DM at Dessie town Hospitals, Northeast Ethiopia, 2020 (n = 322).

Note: OHA: oral hypoglycemic agent, * significant (p < 0.05), ** significant (p < 0.01), *** significant (p < 0.001), 1 - reference, Hosmer - Lemeshow goodness-of-fit (p = 0.781), no multicollinearity (VIF < 10).

7. Discussion

Despite the wide range of effects of T2DM on vision, no thorough study has been undertaken in Ethiopia to describe visual impairment and the factors that influence it among persons with T2DM. The goal of this study was to find out how common visual impairment is among persons with T2DM and what factors contribute to it in Dessie town Hospitals in Northeast Ethiopia.

At Dessie town Hospitals, the prevalence of visual impairment among people with T2DM was 37.58% (95% CI: 32.3–42.9) in the current study. This figure is greater than the 28.9% found in a prior study at Dessie Referral Hospital [41]. This disparity is likely owing to previous researchers’ study designs, which included a review of patient records as a source of data, a different study population (only newly diagnosed DM), and visual disturbance was recognized using clinical findings and questionnaire-based procedures (where visual acuity test was not applied). Furthermore, the current study’s prevalence of visual impairment is higher than other studies in Nigeria (24.1%) [49], Tunisia (22.2%) [39], and Cameron (29.7%). [11] Kumasi, Ghana (18.4%) [37], Zambia (17.1%) [36], Turkey (13.5%) [33], Peru (40.2%) [31], Jordan (17.7%) [32], Hengli, Southern China (10%) [44], and Sankara Nethralaya (4.1%) [30]. This disparity is most likely related to disparities in case definition, technique used, socioeconomic status, and the quality of chronic disease care services provided. The cutoff limit for VI in this investigation was VA 6/12, whereas VA 6/18 was used in the previous trials. Those investigations, unlike the current one, used the better eye’s presenting visual acuity to define visual impairment. Furthermore, their research was conducted at the community level, where there was a chance of screening normal-sighted people. However, Our study was conducted in a hospital setting, and the majority of the patients had a known diabetes problem, which could lead to an increase in the prevalence of visual impairment.

In the Nigerian study, voluntary sampling and a lower sample size were used. Purposive sampling was utilized in Tunisia, which introduced bias, and both investigations defined visual impairment using the better eye presenting visual acuity. When one eye was visually impaired but the other was not, they judged it to be no vision impairment, which understates the extent of visual impairment when compared with the current study, which takes either eye’s visual acuity into account.

This study found a lower rate of unilateral vision impairment than those conducted in Yemen (76.5% [34] and South Africa (78.25%) [35]. This disparity could be attributable to changes in case definitions for vision impairment and sample sizes. In the study conducted in South Africa, study participants were T2DM aged 40, with a cutoff point of VA between 6/9.5 and 6/18, which was defined as a visual impairment, whereas in Yemen, a large sample size was used, with all conditions overestimated or the possibility of adding additional visually impaired cases.

In the current study, visual impairment was significantly associated with advanced age, inadequate regular exercise, diabetes duration, insulin, and poor glycemic management.

For each unit (a year) increase in age, the likelihood of experiencing visual impairment increased by 1.06 times. A study in Tunisia [39], Southern China [16], and Sankara Nethralaya found comparable results [30]. Possible explanations include reduced activity, loss of muscle mass, and weight gain, which cause fatty cells to become more insulin resistant, resulting in hyperglycemia. Due to heart insufficiency, advanced age also increases the risk of macrovascular events [50].

The odds of having visual impairment in diabetes mellitus who had poor physical exercise were 2.91 times more likely than those who were good in physical exercise. This might be due to exercise that can promote an increase in the bioavailability of nitric oxide (NO), which decreases blood pressure, postexercise can increase in glycolipid uptake and utilization, which improves glucose homeostasis, insulin sensitivity and maintaining glycemic level [51, 52, 53], optimized body mass index, and modulated DNA methylation [54].

Participants with duration of diabetes of above 5 years were 2.42 times more likely to get visual impairment as compared with those with type 2 diabetes with duration of 5 years and below. This finding is in line with Zambia [36], Yemen [42], Peru [31], and China [16]. Possible reason might be long duration of diabetes has lower adherence [55], hall marker for long-term exposure to hyperglycemia [56], and potential increase risk of macrovascular and microvascular events and death [50]. Moreover, long duration linked to a reduction in insulin secretion or excessive insulin resistance in T2DM patients [24].

Diabetes mellitus patients who were managed by insulin only were 14 times more likely to have visual impairment than diabetes patients who are managed without insulin. This is consistence with study in Zambia [36], Turkey [33], Peru [31], Jordan [32]. and Sankara Nethralaya [28]. The reason is probably linked to the use of insulin alone that reflects less adherence [55] resulting in deterioration in kidney function, decline in β-cell function, or increase in insulin resistance over time [57], which in turn is associated with poor plasma glucose control and higher risk of severe diabetes.

The odds of being visually impaired were two times higher in poor glycemic control in contrary to good glycemic control, which is in line with study in Peruvians [31]. The possible reason might be poor glycemic control or persistent hyperglycemia damages retinal vasculature via activation of a pro-inflammatory mediators such as tumor necrotic factor (TNF)-2, interleukin-6, interleukin-1b, angiotensin II, endothelin-1, and vascular endothelial growth factor (VEGF) that could alter retinal blood barrier and lead to retinal vessel leakage causing macular edema and nerve scaring, which result in retinal detachment and sudden vision loss.

8. Conclusion

The prevalence of VI in Dessie town hospitals accounts for more than a third of patients living with type-2 DM implied that was a significant public health problem. Older age, poor regular exercise, duration of diabetes, insulin treatment, and poor glycemic control were statistically significant with visual impairment. Regular diabetes follow-up and visual screening for all type-2 DM should be done at older age group patients and for those having longer duration of diabetes, which can reduce visual morbidity and vision loss. Type-2 DM patients should control glycemic level by taking medications and through adequate and regular physical exercise.

The findings of this study are essential for visual health program planners that barriers other than economic constraints are present, which prevent adoption of desired behaviors. Public health policies with educational programs and promotion of DR screening of all T2DM are needed and timely management of DR that greatly reduces the incidence of visual impairment due to diabetes. Thus integrated effort should be in place to reduce the risk of visual impairment, manage the disease progression, and prevent vision loss as a bad consequences.

9. Strength and limitations of the study

9.1 Strength

This study gave a minute picture of the problem (visual impairment) among T2DM patients in Ethiopia and made easy to understand which factors were more important to visual impairment and vision loss as a bad consequence.

9.2 Limitations

Since the study was cross-sectional, it could not show cause-effect relationship. Recall bias was also an expected limitation. Categorization of visual impairment was based on presenting, not corrected visual acuity. HbA1c was not measured due to clients’ financial issue so that physicians ordered fast plasma glucose instead of HbA1c. The HbA1C test is a reliable blood test that provides information about a person’s average levels of blood glucose over the past 3 months. However, the fasting plasma glucose shows only point in time result of the diabetic patients.

10. Recommendation

• Ministry of health should develop effective and efficient health programs, policies, and strategies for preventive, curative, and rehabilitative service specific to diabetics

• Health facilities should have an adequate referral mechanism that allows all patients to screen and diagnose diabetes mellitus early through screening and regular follow-up so that it can detect possible eye complications.

• People living with T2DM should adopt standardized management protocol such as diet, physical activities, and medications to live long and prevent potential complication.

• Researchers should conduct further observational (like cohort) study.