Association of Dyslipidemia with Diabetic Retinopathy in Type 2 Diabetes Mellitus Patients: A Hospital-Based Study.

BACKGROUND: Dyslipidemia is an important risk factor that can lead to the progression of retinopathy (DR). Diabetic dyslipidemia with low high-density lipoprotein (HDL) and increased triglycerides (TGs) are seen frequently among Type 2 diabetic mellitus. AIMS AND
OBJECTIVES: (1) To assess the level of serum lipids (total cholesterol, TGs, HDL, and low-density lipoprotein [LDL]) among type 2 diabetes patients. (2) To determine the association between serum lipid levels and DR.
MATERIALS AND METHODS: This was a hospital-based cross-sectional study conducted in a tertiary care hospital in Salem from September 2018 to March 2020 with a sample size of 200. Details of their diabetic history were obtained. Patients were evaluated for their HbA1C levels, hypertension, and lipid profile status. Early treatment DR Study system was used to classify DR. Low density lipoprotein cholesterol was calculated by Freidewald's equation.
RESULTS: This study showed a significant association among DR and LDL cholesterol. DR with raised LDL, TGs levels, and lowered HDL on adjusted analysis. There was strong association between DR and serum cholesterol in unadjusted analysis; however, there was no association when adjusted for factors such as age, gender, duration of diabetes, and glycemic control. Majority of participants were males (57.5%) with a male: female = 1.35:1. The mean age of the patients in our study was 57.8 (5.8) years and 54.4 (6.6) years in patients with DR and patients without retinopathy, and it was found to be statistically significant. There was a significant difference in the duration of diabetes with the presence of DR and the patients with DR were having longer duration of diabetes (7.9 vs. 6.2 years; P < 0.001). Moderate nonproliferative diabetic retinopathy (NPDR) was found to be present in 41.0% of eyes followed by mild NPDR (20.5% eyes). Proliferative diabetic retinopathy was present only in 9.5%, and the severity of retinopathy was associated only with the HDL level, and there was no association found with total cholesterol, TG, and LDL cholesterol.
CONCLUSION: A statistically significant correlation was found between dyslipidemia and the severity of DR among Type 2 diabetic patients. Copyright:

INTRODUCTION

Diabetes mellitus (DM) has emerged as a leading global health problem, and it is said to be one of the important noncommunicable diseases with considerable morbidity and complications. About 422 million people worldwide have diabetes, the majority are living in low- and middle-income countries, and 1.6 million deaths are directly attributed to diabetes each year.[1] It is estimated to rise to 592 million by 2035.[2] World Health Organization estimates shows that diabetes was the seventh leading cause of death in 2016.[3] Between 2000 and 2016, there was a 5% increase in premature mortality from diabetes.[3]

Vascular abnormalities and microvasculopathy is one of the widely accepted pathogenesis of diabetic retinopathy (DR), and accordingly, the early clinical signs of DR will include microaneurysms and retinal microhemorrhages.[2456]

One of the important risk factor that can be attributed to the progression of DR is dyslipidemia. The classes of lipids affected by these mechanisms are fatty acids, triglycerides (TGs), sphingolipids, and cholesterol.[7] High lipid levels are known to cause endothelial dysfunction due to reduced bioavailability of nitric oxide and this endothelial dysfunction was suggested to play a key role in retinal exudate formation in DR.[8] Hence, the available literature around the globe shows debatable association between serum lipids and DR, with some studies showing that low-density lipoprotein (LDL) and serum total cholesterol are associated with retinopathy while some showing association of TGs with retinopathy.

MATERIALS AND METHODS

Study setting

This study was conducted at the Department of Ophthalmology at Vinayaka Mission's Kirupananda Variyar Medical College and Hospital for a period from September 2018 to March 2020.

Study population

Diabetic patients referred for the evaluation of diabetic retinopathy and all patients with Type 2 DM attending Outpatient Department of Vinayaga Mission Medical College and Hospital were taken in this study.

Study design

This is a descriptive cross sectional study with a sample size.

Inclusion criteria

Patients with Type 2 DM for more than 3 years, aged 45 years and above attending the Department of Ophthalmology, Vinayaka Mission's Kirupananda Variyar Medical College and Hospital, Salem, were included in the study. They were divided into those with retinopathy and those without retinopathy.

Exclusion criteria

Patients with duration of diabetes <3 years and other ocular comorbidities. Patients with hypertension, cardiovascular diseases, renal and liver disorders, severe anemia, and thyroid disorders are excluded. Patient on hypolipidemic drugs and who have been treated earlier with either LASER or intravitreal antivascular endothelial growth factor injections are excluded.

Sampling method

DR patients meeting the inclusion and exclusion criteria were selected using convenient sampling method, from the available records maintained in the department.

Data collection

Diabetic patients 100 samples who met all inclusion and exclusion criteria were chosen and relevant history regarding the patient's diabetes was taken age of onset of diabetes [first diagnosed], duration of diabetes, nature and duration of treatment received, history regarding patient's glycemic control, and compliance to the treatment. Patient's examination was then performed as per the pro forma. The blood investigations such as fasting and postprandial sugar, HbA1C, serum total cholesterol, TGs, and high-density lipoprotein (HDL) cholesterol were estimated by enzymatic methods. LDL cholesterol was calculated by Friedewald's equation.

Statistical analysis plan

Data were analyzed using SPSS (Statistical package for social sciences) V21 f or windows. Categorical variables such as gender; type of treatment is expressed as frequency and percentages. Continuous variables such as age, duration of diabetes, and lipid levels are expressed as mean (standard deviation) or median (interquartile range), depending on the type of distribution. Chi-square test was used to determine the association between the categorical variables. Independent sample t-test was used to compare the serum lipid levels between DR and without DR among Type 2 diabetes. One-way ANOVA with Bonferroni post hoc test was used to determine the association between the severity of retinopathy and the duration of diabetes. A P < 0.05 was considered statistically significant.

Approval of research review board

The ethical approval was sought from the Institutional Ethics Committee. Informed consent was obtained from the study participants before data collection. Confidentiality was maintained by limiting the identifying variables to the minimum. Data were analyzed in aggregate, and access to the collected data was limited only to me, my guide, and co-guide.

RESULTS

Table 1 compares the baseline characteristics of gender and age between the two groups. It shows that the males and females are comparable between the two groups. The mean age of the patients with DR was 57.8 years, and the mean age of the patients without DR was 54.4 years, and it was found to be statistically significant (P < 0.001).

Table 1

Sociodemographic characteristics of the participants (n=200)

Characteristics	Diabetic cases		P
	With retinopathy (n=100)	Without retinopathy (n=100)
Males	63 (63.0)	52 (52.0)	0.116
Females	37 (37.0)	48 (48.0)
Age (years), mean (SD)	57.8 (5.8)	54.4 (6.6)	<0.001*

*Independent sample. SD: Standard deviation

Table 2 compares the clinical profile of the diabetic patients with and without retinopathy. There was a significant difference in duration of diabetes between the patients with DR and those without retinopathy. The patients with retinopathy had significantly longer duration of diabetes at the time of presentation to the hospital (7.9 vs. 6.2 years: P <0.001). The type of treatment was comparable between the groups.

Table 2

Clinical profile of the study participants (n=200)

	Diabetic cases		P
	Without retinopathy (n=100)	With retinopathy (n=100)
Duration of DM (years), mean (SD)	6.2 (1.3)	7.9 (3.7)	<0.001*
Type of treatment, n (%)
Oral hypoglycemic drugs	88 (88.0)	77 (77.0)	0.234
Native treatment	12 (12.0)	23 (23.0)

*Independent samples t-test. SD: Standard deviation, DM: Doctorate of medicine

Table 3 shows the severity of retinopathy and the presence of clinically significant macular edema among the patients with DR. Moderate nonproliferative diabetic retinopathy (NPDR) was found to be present in 41.0% eyes followed by mild NPDR (20.5% eyes). Proliferative diabetic retinopathy (PDR) was present only in 9.5%.

Table 3

Distribution of participants by severity of retinopathy and presence of clinically significant macular edema (n=200)

Severity of retinopathy	n (%)	CSME, n (%)
Mild NPDR	41 (20.5)	-
Moderate NPDR	82 (41.0)	20 (24.4)
Severe NPDR	38 (19.0)	8 (21.0)
Very severe NPDR	20 (10.0)	2 (10.0)
PDR	19 (9.5)	6 (31.6)

NPDR: Nonproliferative diabetic retinopathy, PDR: Proliferative diabetic retinopathy, CSME: Clinically significant macular edema

Table 4 shows the association of lipid levels with DR. It can be noted that mean total cholesterol, TGs, and LDL were significantly higher in patients with retinopathy when compared to patients without retinopathy (P < 0.05). There was an inverse relationship with HDL cholesterol, and it was found to be significant. When considered separately, the mean lipid levels (total cholesterol, TGs, and LDL) were high and points toward dyslipidemia. HDL cholesterol was very low in patients with DR.

Table 4

Association between diabetic retinopathy and serum lipid levels (n=200)

	Diabetes mellitus (mg/dl), mean (SD)		P
	Without retinopathy (n=100)	With retinopathy (n=100)
Total cholesterol	219.5 (26.5)	228.6 (30.9)	0.027
Triglycerides	160.6 (9.8)	179.9 (20.5)	<0.001
LDL	115.7 (8.1)	131.9 (16.4)	<0.001
HDL	47.1 (10.8)	38.5 (3.1)	<0.001

SD: Standard deviation, LDL: Low-density lipoprotein, HDL: High-density lipoprotein

Table 5 shows that dyslipidemia was more common among the patients with DR when compared to patients without retinopathy. The difference was significant for total cholesterol, TGs, and HDL.

Table 5

Classification of the participants according to dyslipidemia (n=200)

	Diabetes mellitus
	With retinopathy (n=100), n (%)	Without retinopathy (n=100), n (%)
Total cholesterol (mg/dl)*
Normal	16 (16.0)	29 (29.0)
High	84 (84.0)	71 (71.0)
Triglycerides (mg/dl)*
Normal	4 (4.0)	17 (17.0)
High	96 (96.0)	83 (83.0)
LDL (mg/dl)
Normal	2 (2.0)	0
High	98 (98.0)	100
HDL (mg/dl)*
Normal	12 (12.0)	76 (76.0)
Low	88 (88.0)	24 (24.0)

*Significant difference between the two groups (P<0.05). LDL: Low-density lipoprotein, HDL: High-density lipoprotein

Table 6 shows that, although the mean LDL was found to higher among those patients with proliferative DR, it was not found to be statistically significant.

Table 6

Association of low-density lipoprotein cholesterol with the severity of diabetic retinopathy (n=200)

Severity of retinopathy	n	LDL cholesterol (mg/dl), mean (SD)	P*
Mild NPDR	41	130.8 (17.1)	0.294
Moderate NPDR	82	130.8 (15.8)
Severe NPDR	38	133.9 (16.0)
Very severe NPDR	20	128.8 (14.8)
PDR	19	138.5 (18.4)

*One way ANOVA. SD: Standard deviation, LDL: Low-density lipoprotein, NPDR: Nonproliferative diabetic retinopathy, PDR: Proliferative diabetic retinopathy

Table 7 shows the association between mean total cholesterol and DR after adjusting for confounders using multivariable linear regression. It can be seen that retinopathy is associated with increased total cholesterol on unadjusted analysis (r = 9.1 [95% confidence interval: 1.1–17.1, P = 0.027]), but there was no significance when adjusted for age, gender, duration of diabetes, and the control status.

Table 7

Univariate and multivariable linear regression for association of total cholesterol with risk factors (n=200)

	Unadjusted r (95% CI)	P	Adjusted r (95% CI)	P
Diabetic retinopathy	9.1 (1.1-17.1)	0.027	9.0 (−1.1-19.1)	0.080
Age (years)	0.3 (−0.3-0.9)	0.834	0.2 (−0.5-1.0)	0.523
Male gender	0.3 (−7.9-8.6)	0.933	1.3 (−7.1-9.8)	0.754
Duration of diabetes (years)	0.2 (−1.2-1.6)	0.748	−0.4 (−2.1-1.2)	0.346
HbA1C (mg/dl)	−1.1 (−3.7-1.4)	0.384	−0.2 (−0.1-0.1)	0.903
FBS (mg/dl)	0.04 (−0.09-0.17)	0.533	−0.001 (−0.24-0.24)	0.991
PPBS (mg/dl)	0.02 (−0.04-0.08)	0.503	−0.01 (−0.11-0.11)	0.992

CI: Confidence interval, FBS: Fasting blood sugar, PPBS: Postprandial blood sugar, HbA1C: Glycated hemoglobin

Table 8 shows that there was a significant positive correlation for DR, age, duration of diabetes, fasting blood glucose (FBS), and postprandial blood sugar (PPBS) with LDL cholesterol on univariate analysis. However, on multivariable analysis, the significance was present only for DR.

Table 8

Univariate and multivariable linear regression for association of low-density lipoprotein cholesterol with risk factors (n=200)

	Unadjusted r (95% CI)	P	Adjusted r (95% CI)	P
Diabetic retinopathy	16.2 (12.6-19.8)	<0.001	15.3 (10.8-19.0)	<0.001
Age (years)	0.5 (0.2-0.8)	0.003	0.1 (−0.2-0.5)	0.473
Male gender	−0.3 (−4.6-3.9)	0.886	2.0 (−1.8-5.8)	0.297
Duration of diabetes (years)	1.1 (0.4-1.8)	0.002	0.3 (−0.4-1.0)	0.647
HbA1C (mg/dl)	−1.8 (−3.2-−0.5)	0.007	−0.3 (−1.6-1.0)	0.647
FBS (mg/dl)	0.09 (0.02-0.16)	0.007	0.03 (−0.07-0.14)	0.521
PPBS (mg/dl)	0.04 (0.01-0.07)	0.010	−0.02 (−0.06-0.03)	0.509

CI: Confidence interval, FBS: Fasting blood sugar, PPBS: Postprandial blood sugar, HbA1C: Glycated hemoglobin

DISCUSSION

Our study reported that majority of the participants were males (57.5%) with a male: female = 1.35:1. There was a higher percentage of males in group with DR when compared to group without DR, but they were comparable (P = 0.116). The results are similar to the studies conducted by Hegde and Vekategowda,[8] Rema et al.,[9] and Paromita et al.[10] where male preponderance was reported, and it was comparable between the groups. A study by Mohan et al.[11] in a clinical cohort in Chennai on DR, the male preponderance was present and male: female = 2:1. Similarly, a study by showed high male preponderance.

The mean age of the patients in our study was 57.8 (5.8)–54.4 (6.6) years in patients with DR and patients without retinopathy, and it was found to be statistically significant. Literatures have shown that the risk of DR is higher after 10 years of diagnosis of diabetes and hence it is always possible to get an association. This finding can be more confirmed by results of other studies conducted across the countries which has shown that there was a significant association between increasing age and DR prevalence.[11]

The classes of lipids affected by these mechanisms are fatty acids, TGs, sphingolipids, and cholesterol. Above all these, insulin resistance by itself results in raised LDL, TG level, total cholesterol, and free fatty acids which results in dyslipidemia. It also leads on to decrease in HDL cholesterol, and also reverse cholesterol transport genes are also inhibited by insulin resistance.[12]

Dornan et al.[13] were the first to show an association between DR and LDL cholesterol in his landmark study. A study by Rema et al.[10] showed a similar result as our study where total serum cholesterol, LSL, TG, and HDL were associated with DR on unadjusted analysis, but the significance was lost for total cholesterol when adjusted for age, gender, duration of diabetes, and control status. Many other studies showed a significant association with dyslipidemia.[14] A study by Sinav et al.[15] and Larsson et al.[16] showed a significant association with total cholesterol, LDL, and HDL, but the association was absent for TGs. However, a study by Hove et al.[17] and Hegde and Vekategowda[9] showed a significant association with all the parameters.

HDL was significantly associated with the severity of DR. The postulate is that dyslipidemia causes increased viscosity and alterations in the fibrinolytic system which forms hard exudates. In addition to it, when TG gets incorporated into the cell membrane, there is change in the membrane fluidity and leakage of plasma and its constituents into the plasma. This in turn results in retinal hemorrhage. In addition to that, it also causes endothelial dysfunction which also worsens the retinopathy.

Strengths and limitations

The main strength of the study is the inclusion of large representative sample of diabetic patients in our study which resulted in the increased power of our study, and hence the generalizability to the similar setting. Dedicated software was used for the data collection and analysis which had reduced the chance of errors. One of the limitations is that, since the study was a hospital-based study, there was a chance of selection bias as more cases with uncontrolled diabetes would have been referred to the hospital which would have affected the relationships. As with all other cross-sectional studies, temporal association could not be established with this study, since there is always a chance of reverse causal association and hence a cohort study with a long follow-up would have been more ideal.

CONCLUSION AND RECOMMENDATIONS

Majority of the participants were males (57.5%) with a male: female = 1.35:1. There was a higher percentage of males in group with DR when compared to group without DR, but they were comparable (P = 0.116). The mean age of the patients in our study was 57.8 (5.8) years–54.4 (6.6) years in patients with DR and patients without retinopathy, and it was found to be statistically significant. There was a significant difference in the duration of diabetes with the presence of DR, and the patients with DR were having longer duration of diabetes (7.9 vs. 6.2 years; P < 0.001). Visual impairment was more among patients with DR, and it was found to be statistically significant. DR with raised LDL, TGs, and lowered HDL on adjusted analysis. There was an association between DR and serum total cholesterol in unadjusted analysis; however, there was no association when adjusted for age, gender, duration of diabetes, and the glycemic control as represented by the HbA1C, FBS, and PPBS levels. Moderate NPDR was found to be present in 41.0% of eyes followed by mild NPDR (20.5% eyes). PDR was present only in 9.5%, and the severity of retinopathy was associated only with the HDL level and there was no association with total cholesterol, TG, and LDL cholesterol. Hence, it is recommended that more strict measure has to be taken to control the modifiable risk factors associated with the progression of DR and hence the prevention of visual loss. These findings open our eyes broader since there is still discrepancy in the findings on whether DR is associated with dyslipidemia, and this exactly explains the reason behind differing hyperlipidemia therapies among the retinopathy patients.

Data availability statement

The database generated and analyzed during this study is available from the corresponding author on reasonable request.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.